University Micrdnlms International 7908232

INFORMATION TO USERS

This was produced from a copy of a document sent to us for microfilming. While the most advanced technological means to photograph and reproduce this document have been used, the quality is heavily dependent upon the quality of the material submitted.

The following explanation of techniques is provided to help you understand markings or notations which may appear on this reproduction.

1. The sign or “target” for pages apparently lacking from the document photographed is “Missing Page(s)”. If it was possible to obtain the missing page(s) or section, they are spliced into the fHm along with adjacent pages. This may have necessitated cutting through an image and duplicating adjacent pages to assure you of complete continuity.

2. When an image on the film is obliterated with a round black mark it is an indication that the film inspector noticed either blurred copy because of movement during exposure, or duplicate copy. Unless we meant to delete copyrighted materials that should not have been filmed, you will find a good image of the page in the adjacent frame.

3. When a map, drawing or chart, etc., is part of the material being photo graphed the photographer has followed a definite method in “sectioning” the material. It is customary to begin filming at the upper left hand comer of a large sheet and to continue from left to right in equal sections with small overlaps. If necessary, sectioning is continued again—beginning below the first row and continuing on until complete.

4. For any illustrations that cannot be reproduced satisfactorily by xerography, photographic prints can be purchased at additional cost and tipped into your xerographic copy. Requests can be made to our Dissertations Customer Services Department.

5. Some pages in any document may have indistinct print. In all cases we have filmed the best available copy.

University Micrdnlms International 7908232

WANG, YING STUDY OF THE FORMATION AND REACTIVITY OF ORGANIC CATIONS IN SOLUTION BY THE PULSE RADIOLYSIS METHOD.

THE OHIO STATE U N IV E R S IT Y , P H .D ., 1978

International 300 n. zees roao. ann a rb o r, mi abioe STUDY OF THE FORMATION AND REACTIVITY OF ORGANIC

CATIONS IN SOLUTION BY THE PULSE RADIOLYSIS METHOD

DISSERTATION

Presented in Partial Fulfillment of the Requirements for

the Degree Doctor of Philosophy in the Graduate

School of The Ohio State University

Ying Wang, B.S., M.S.

The Ohio State University

1978

Reading Committee; Approved hy

Richard F. Firestone

Jack Hine

Leon M. Dorfman Adviser Department of Chemistry TO MY PARENTS ACKNOWLEDGMENT

I wish to thank my adviser, Professor Leon M. Dorfman, for his constructive comments and constant encouragement throughout the course of this investigation. I also wish to express my appreciation to Professor Jack Hine for helpful discussion on the subject of structure-reactivity relationship and to Mr. Ed Ray for maintaining the electronic equipment. I am grateful to the Department of Chemistry of The

Ohio State University and to the United States Department of Energy for their financial support.

Finally, I wish to thank my wife, Shu-Han, for her patience, understanding, and support throughout the course of this work. June 2, 1950 --- B o m - Taiwan, China

1968-1972...... B.S., National Chung-Hsing University,' Taiwan

1972-1974...... Military service

1974-1976 .... Teaching Associate, Department of Chemistry, The Ohio State University, Columbus, Ohio

1974-1976 .... M.S., The Ohio State University, Columbus,,Ohio

.1976-1978....— Research Associate, Department of Chemistry, The Ohio State University, Columbus, Ohio

PUBLICATIONS

(1) "The Formation of Carbénium Ions, Carbanions and Carbanioh-Pairs in Irradiated Solutions, and the Reactivity of These Intermediates" Leon M. Dorfman, Ying Wang, Hsien-Yien Wang, and Richard J. Sujdak, Faraday Discussion No. 6 3, radiation effects-63/l3(l976).

(2) "Fast Reaction Studies of Carbanions and Carbocations in Solution" Leon M. Dorfman, Vincent M. DePalma and Ying Wang, in "Protons and Ions Involved in Fast Dynamic Phenomena", Elsevier Scientific Publishing Co., 1978, Proceedings of 30th Meeting of Société de Chimie Physique, Paris, November 1977. (3) "The Reactivity of Arylcarbénium Ions with Alkylamines and Ammonia in Solution: Kinetic Effect of Cyclopropyl Substitution; Solvent Effect" Vincent M. DePalma, Ying Wang, and Leon M. Dorfman, J. Amer. Chem. Soc., 100, 5416(1978)

(4) "The Reactivity of Arylcarbenium Ions with Alkenes", Ying Wang and Leon M. Dorfman, in preparation.

(5) "Formation Mechanism of Arylcarbénium Ions in the Irradiated 1 ,2-Dichloroethane solution", Ying Wang, John Tria, and Leon M. Dorfman, in preparation. TABLE OF CONTENTS Page

ACKNOWLEDGMENT ...... ii

VITA ...... iii

LIST OF TABLES ...... vi LIST OF FIGURES ...... vii

LIST OF SYMBOLS AND ABBREVIATIONS ...... xiii

INTRODUCTION ...... 1

I. Carbénium Ions ...... 1 II. Carbénium Ions Formed by Ionization Radiation...3

EXPERIMENTAL ...... 12

. I. Pulse Irradiation ...... 12

II. Optical Detection ...... l4

III. Reaction Cells ...... 22

IV. Sample Preparation ...... , 28

V. Dosimetry...... 33

VI. Data Analysis ...... 33

VII. Materials ...... 36

RESULTS AND DISCUSSION ..... 40

I. Mechanism of the Formation of Arylcarbénium Ions in the Irradiated 1 ,2-Dichloroethane Solution ...... 40 Yield Curve Studies ...... 4l

Competition Kinetics ...... 6l

Titration Experiments ...... 82

Proposed Mechanism ...... 95 iv TABLE OF CONTENTS (continued)

II. Reactions Between Arylcarbenium Ions and Alkenes ...... IO9

Possible Reaction Path ...... 112

Ethylene, Propylene, Isobutylene, and Cyclohexene ...... II6

Ethylene, Propylene, Neohexene, and Allylbenzene ...... 129

Propylene, 1,3-Butadiene, Cyclohexene, and 1»3-Cyclohexadiene ...... 136

1 .3-Butadiene, 2-Methyl-1,3-butadiene, 2.3-Dimethyl-l,3-butadiene, 4-Methyl-l,3- pen tadiene, and 2,4-Dimethyl-l,3-pentadiene..146

APPENDIXES ...... 161

A. Competition Kinetics - Two Solutes Competing for One Solvent Cation ...... I6I

B. Competition Kinetics - Two Solutes Competing for Two Solvent Cations ...... I63

C. Pseudo-First Order Rate Equation with Overlapping Absorption ...... I66

D. Calculations of AH's for Reactions between Benzyl Cation and Alkenes ...... l6?

REFERENCES AND NOTES ...... I69 LIST OF TABLES

Table Page

1. Solubilities of alkene gases in 1,2-DCE at 25°C ...... 32 2. Rate constants between several carbénium ions and ammonia at 24 C in 1,2-DCE ..... 87 3. Propagation rate constants for the cationic polymerization of styrene ...... Ill

4. Rate constants for reactions between benzyl cation and alkenes in 1,2-DCE at 24^C...... 122

5. Gas phase carbénium ion affinities of olefins...... 12?

6. Rate constants for reactions between benzyl cation and alkenes in 1,2-DCE at 24°C...... 134

7. Rate constants for reactions between benzyl cation and alkenes in 1,2-DCE at 24°C ...... 143

8. Rate constants for reactions between benzhydryl cation and 1.3-conjugated dienes in 1,2-DCE at 24°C ...... 156

9. Calculations of AH's for reactions between benzyl cation and alkenes in gas phase ...... l68

10. Summary of the rate constants for the reactions between arylcarbenium ions and alkenes in 1,2-DCE at 24 C ...... l60 LIST OF FIGURES

Figure Page

1. Schematic diagram of the inhomogeneous distribution of the primary species formed in the irradiated liquid...... 6

2. The optical absorption spectra of benzyl cation, benzhydryl cation, and trityl cation ...... 10

3. Time profile of an 85 ns pulse and a 600 ns pulse ...... 15

4. Schematic diagrams of the double pass optical detection system and of the single pass detection system...... 17

5. Spectral radiant energy distribution for Rudolph xenon lamp with 450 watt bulb .... 19

6. Time profile of xenon flash at 405 nm .... 20

7. Spectral responses of phototubes with s-1, s-5 , and s-19 response ...... 23

8. Reaction cells used in pulse radiolysis experiments ...... 25

9. A cryostat used in temperature dependent experiment ...... 27

10. Apparatus used for the determination of solubility of gas in 1 ,2-dichloroethane solution ...... 30

11. Two typical oscilloscope traces showing formation following the pulse and decay following the pulse ...... 34

12. The plot of optical density at 364 nm vs. the concentration of dibenzylmercury in the irradiated dibenzylmercury-1,2-DCE solution..44

1 3. The time dependent spectrum resulting from irradiation of a O.OO55 M solution of dibenzylmercury in 1,2-DCE under vacuum.... 45 vii LIST OF FIGURES (continued)

Figure Page

14. The time dependent spectrum of irradiated propylene-dihenzylmercury-l,2-DCE solution under vacuum ...... 46

1 5. The decay curve of "benzyl cation at 369 nm in the presence of 0.742 M propylene ..... 48

1 6. The plot of 1/0 .D. of benzyl cation at 364 nm as a function of time in the irradiated dibenzylmercury-1,2-DCE solution ...... 49

1 7. The plot of optical density taken immediately after the pulse at 449 nm vs. the concentration of benzhydryl bromide... 51

1 8. The plot of optical density at 449 nm vs. the concentration of diphenylmethanol in the irradiated diphenylmethanol-1,2-DCE solution ...... 53

1 9. The rate curve of benzhydryl cation at 449 nm in the irradiated diphenylmethanol-DCE solution ...... 54

20. The plot of optical density at 434 nm as a function of the concentration of triphenylmethanol in the irradiated triphenylmethanol-DCE solution ...... 55

21. The rate curves of the irradiated triphenylmethanol-DCE solution at 434 nm and at 506 n m ...... 56

22. The plot of the optical density of trityl cation at 439 nm vs. the concentration of triphenylmethyl bromide in 1,2-DCE ...... 58

2 3. A summary of the yield curve studies...— 59

24. The decay curves of benzyl cation in the irradiated ammonia-dibenzylmercury-DCE solution for two different ammonia concentrations ...... 64 LIST OF FIGURES (continued)

Figure Page

2 5. The plot of -In O.D. of "benzyl cation VS; time in the irradiated ammonia- dihenzylmercury-DCE solution...... 66

2 6. The plot of (0D/0D*)-1 v s . [ammonia] / [dibenzylmercury] .... *...... 6?

2 7. The decay curve of trityl cation at 434 nm in the irradiated ammonia-triphenyl methanol-DCE solution ...... 69

28. The plot of (0D/0D*)-1 vs. [ammonia]/ [triphenylmethanol]...... 70

2 9. The decay curves of benzhydryl cation in the irradiated ammonia-benzhydryl bromide- DOE solution for two different ammonia concentrations...... 72

3 0 . The plot of (0D/0D*)-1 vs. [ammonia]/ [benzhydryl bromide] ...... 73

3 1. The plot of OD*/OD vs. [ammonia]/ t»enzhydryl bromide] ...... 76

3 2. The plot of (OD/OD*)-1 vs. [ammonia] / [anthracene] ...... 80

33* The time dependent spectrum resulting from irradiation of a 0 .0 3 9 M solution of 1 ,1-diphenylethylene in 1,2-DCE under 1 atm air 85

3 4. Comparison of the optical absorption spectra of the substituted benzhydryl cations from pulse radiolysis experiment and from literature ..... 86 35* The plot of pseudo-first order rate constant vs. the concentration of ammonia for the reaction between substituted benzhydryl cation and ammonia in 1,2-DCE at 24 C 88 LIST OF FIGURES (continued)

Figure Page

3 6. The plot of optical density at 4-30 nm vs. the concentration of diphenylethylene in the irradiated diphenylethylene-DCE solution...... 89 3 7. The decay curve of substituted benzhydryl cation at 4-30 nm in the irradiated diphenylethylene-DGE solution ...... 90

3 8. The time dependent spectrum resulting from irradiation of a 0.304- M solution of DMPD in 1,2-DCE ...... 93

3 9 . The time dependent spectrum resulting from irradiation of a 0.304- M solution of DMPD in 1,2-DCE saturated withoxy g e n...... 94-

4-0. The plot of pseudo-first order rate constant vs. the concentration of ammonia for the reaction between tetrasubstituted allyl cation and ammonia in 1,2-DCE ...... 96

4-1. The formation rate curve of p-terphenyl radical cation monitored at 96O n m ...... I05

4-2. The formation rate curve of benzhydryl cation monitored at 4-4-9 nm ...... IO6

4-3. The time dependent spectrum resulting ■ from irradiation of an isobutylene- dibenzylmercury-DCE solution under vacuum.. 117

4-4-. The decay curves of benzyl cation in the presence of 0.097 M isobutylene and in the presence of O .055 M cyclohexene ...... 118

4-5. The plot of pseudo-first order rate constant vs. the concentration of propene for the reaction between benzyl cation and propene in 1,2-DCE ...... 119 LIST OF FIGURES (continued)

FigLire Page

46. The plot of pseudo-first order rate constant vs. the concentration of isobutylene for the reaction between benzyl cation and isobutylene ...... 120

4 7. The plot of pseudo-first order rate constant vs. the concentration of 1,3-butadiene for the reaction between benzyl cation and 1,3-butadiene ...... 121

48. The plot of log k vs. AH for reactions between benzyl cation and alkenes ...... 125

4 9. The plot of log k vs. AH for reactions between carbénium ions and alkenes ..... 128

5 0. The time dependent spectrum resulting from irradiation of a neohexene- dibenzylmercury-DCE solution ...... I30

5 1. The time dependent spectrum resulting from irradiation of an allyl benzene- dibenzylmercury-DCE solution ...... 13I

5 2. The plot of pseudo-first order rate constant vs. concentration of neohexene for the reaction between benzyl cation and neohexene ...... 132 5 3. The plot of pseudo-first order rate constant vs. concentration of allyl benzene for the reaction between benzyl cation and allyl benzene ...... 133 5 4. The time dependent spectrum resulting from irradiation of a 1 ,3-butadiene- dibenzylmercury-DCE solution ...... 137

55* The time dependent spectrum resulting from irradiation of a 1 ,3-cyclohexadiene- dibenzylmercury-DCE solution ...... 138 LIST OF FIGURES (continued)

Figure Page

56. The decay curve of benzyl cation at 364 nm in the presence of 0.72? M 1,3- butadiene ...... l40

5 7. The plot of pseudo-first order rate constant vs. concentration of cyclohexene for the reaction between benzyl cation and cyclohexene in 1,2-DCE ...... l4l

5 8. The plot of pseudo-first order rate constant vs. concentration of 1,3- cyclohexadiene for the reaction between benzyl cationand 1,3-cyclohexadiene ..... 142

5 9. The time dependent spectrum resulting from irradiation of DMPD-benzhydryl bromide- DCE solution ...... 147

6 0. The time dependent spectrum resulting from irradiation of a MPD-benzhydryl bromide-DCE solution ...... 150

6 1. The optical absorption spectra resulting from irradiation of a 0.27 M solution of MPD in DOE under vacuum and 0.20 M solution of MPD in DOE in the presence of 0.002 M of ammonia ...... 151 6 2. The plots of pseudo-first order rate constants vs. concentrations of isoprene, DMBD, and MPD...... 154

6 3. The plot of pseudo-first order rate constant vs. concentration of DMPD...... 155

64. The formation rate curve of tetrasubstituted allyl cation at 312nm ...... 157 LIST OF SYMBOLS AND ABBREVIATIONS

SIMBOLS;

@ - cartenium ion

t - radical cation + - positive charged species

ABBREVIATIONS ;

1,2-DCE - 1,2-dichloroethane

Am - aromatic hydrocarbon BD - 1,3-butadiene

MBD - 2-methyl-l,3-butadiene

DMBD - 2,3-dimethyl-1,3-butadiene

MPD - 4-methyl-l,3-pentadiene

DMPD - 2,^-dimethyl-1,3-pentadiene PhCH® - phenylcarbenium ion (benzyl cation)

Ph2CH® - diphenylcarbénium ion (benzhydryl cation)

Ph^G® - triphenylcarbénium ion (trityl cation) INTRODUCTION

CARBENIUM IONS

Organic species that contain a carbon atom sharing only six electrons are referred to as carbénium ions. Triphenyl- carbénium ion was the first carbénium ion to be studied and identified during the early twentieth century^0,11,12^ Since then carbénium ions have been recognized as important organic intermediates in many reactions such as solvolyses, Friedel-

Crafts reactions, cationic polymerizations, molecular rearrangements, and hydride transfer reactions. As a matter of fact, the idea that carbénium ions might appear as intermediates in the course of reactions that start from non-ionic reactants and lead to non-ionic products has proved to be one of the boldest and most fruitful concepts born of the 1920

Although stable solutions of triarylcarbénium ions could be prepared and their crystalline salts is''.ated during the

1 9 2 0's, attempts to obtain stable solutions of carbénium ions other than triarylcarbénium ions have been undertaken only since 1950 due to the high reactivity of these carbénium ions toward various nucleophiles. Since investigations made in the decades 193O-I950 suggested the presence of highly reactive carbénium ions as intermediates, it was natural to

1 2

try to confirm their existence by isolation as stable

salts or preparation of stable solutions of the carbénium

ions. Notable success in this field was achieved by Olah

and co-workers. By using the so called super acid system .

(FSO^H-SbF^), they could prepare stable solutions of carbénium

ions by the following simple reaction^-^;

(1)

Many carbénium ions such as alkyl carbénium ions can now be generated in stable form and their properties can be studied.

Through many decades of research carbénium ion chemistry has now been developed into one of the most fruitful fields in physical organic chemistry. Various techniques have been used to investigate the properties of carbénium ions include cryoscopie measurements, conductivity measurements, electronic spectra, vibrational spectra, nuclear magnetic resonance spectra, mass spectrometry, ion cyclotron resonance spectroscopy, stopped flow method, and solution calorimetry^^.

A monograph of five volumes dealing with the chemistry of carbénium ions has been published^"^.

However, the direct kinetic study of these carbénium ions is still limited to those relatively stable ones such as triarylcarbénium ions^®’^^. Available experimental techniques such as spectrophotometric and stopped flow methods with time resolution of greater than millisecond restricted the measurement of absolute reactivity of less stable carbénium ions. Consequently, much of the current knowledge concerning their chemical reactivity has resulted from indirect kinetic study such as competition experiments^^, which indicate only relative reactivity.

By using the fast reaction technique of pulse radiolysis it is now possible to generate reactive carbénium ions in solution on a submicrosecond time scale and to determine their optical absorption spectra^^. Because of the time resolution this method affords, absolute rate constants for many of their elementary reactions can be determined^^”^^.

It is appropriate at this point to briefly indicate how carbénium ions are formed by the radiation-chemical method.

CARBENIUM IONS FORMED BY IONIZATION RADIATION

Some knowledge of the processes by which radiation interacts with matter is essential to an understanding of radiation-chemical phenomena, since the chemical effects are a direct consequence of the absorption of energy from the radiation. Only a brief introduction will be presented here; detailed discussion can be obtained elsewhere^^’^"^.

Electrons interact with matter by a number of processes of which the most important are; (1) the emission of electromagnetic radiation, (2) inelastic collisions, and (3) elastic scattering^^. The relative importance of these

processes depends strongly on the energy of the incident

electrons and, to a smaller extent, on the nature of the

absorbing material. At an electron energy between 10 and 100

Mev (the exact energy depends on the stopping material), energy

is lost predominantly by radiation emission. The emission

energy spectrum extends from zero to the energy of the incident

electrons. At very low electron energy (about several hundred

Kev), elastic scattering is of greatest importance. In this

process, electrons are deflected by the Coulomb field of an

atomic nucleus without conversion of kinetic energy to any other form of energy. For electrons with medium amount of

energy such as 4 Mev electrons (initially) used in this work,

inelastic collisions are the dominant process. Energy is

lost through Coulomb interactions with electrons of the

stopping material. When the moving electron approaches a molecule lying near its path, a fraction of its energy is

transferred through Coulomb interactions to the electrons of

that molecule and as a result these electrons are either

expelled (ionization) or displaced from their ground states

(excitation). Electrons ejected as a consequence of the

ionization produced by radiation may have a kinetic energy which is sufficient to produce further ionization and

excitation. If the energy of these secondary electrons is less than about 100 ev, their range in liquid will be short and any secondary ionizations that they produce will be situated close to the original ionization, giving a "spur" of excited and ionized species. For secondary electrons in the energy range 100 to 500 ev, a large spur is produced called a "blob". Higher energy secondary electrons will be sufficiently energetic to travel further from the site of the original ionization and will form tracks of their own, branching off from the primary track. This is shown in

Figure 1.

The above processes can be summarized in the following equation:

S (S*. S*. e-) + e- (2)

where S represents a solvent molecule. The ions, excited molecules, and electrons are enclosed in brackets as a reminder that they are largely grouped in spurs, blobs, and tracks. A portion of the secondary electrons are ejected with sufficient energy to create tracks of their own and inevitably become widely separated from their parent ions; these are shown as isolated electrons in equation (2).

Subsequent processes in the spurs, blobs, and tracks may include ion dissociation, ion-molecule reaction, geminate recombination of ions, dissociation of excited molecules to radicals or molecular products, recombination of caged radicals, etc. The ejected electrons are rapidly thermalized, and, in a polar medium, the molecules of the blobs

100 to 500 ev spurs primary excited species

>5000 ev short tracks

branch tracks

Figure 1 . Schematic diagram of the inhomogeneous distribution of the primary species formed in the irradiated liquid. 7 liquid become oriented around them, forming so-called

"solvated electrons". Sometimes the electrons are attached to the solvent molecules through non-dissociative or dissociative processes, forming anionic species. These in-spur reactions, electron attachment and solvation processes usually occur in times between 10“^^ and 10“^ seconds after passage of the electrons^^'^"^. As time increases, the spurs, blobs, and tracks will expand by diffusion until ultimately a homogeneous distribution is set up. The intermediates which "escape" the spur to react homogeneously in the bulk of the liquid are either ions

("free ions"), radicals, or long-lived excited molecules.

It is mainly these "free" species that enable us to control the chemistry of the irradiated system by adding an appropriate solute to react with. In the present discussion, attention will be focused on free ions only.

Under appropriate circumstances, the electron-a reducing species-and the positively charged solvent ion-an oxidizing species-may react with the solute to produce radical ions.

By selecting the proper chemical systems, only positive ions or negative ions will be produced. For example, in alcohol solvents, various aromatic radical anions have been produced by radiolysis^^,29,30 ^ The primary species ROht abstracts a proton from the solvent to form ROH^ which does not ionize most aromatic solutes^^. However, the electrons are solvated by alcohol molecules and remain free to react with aromatic molecules.

^sol Am ---- > Am~ (3) where Am represents the aromatic solute.

If a chlorocarhon such as 1,2-dichloroethane (1,2-DCE) is used as solvent, positively charged solute ions will he formed.- The electron formed by radiation reacts with a

1,2-DCE molecule through dissociative attachment to form 21 a free radical and a halide ion :

(4)

Thus, the electron, a transient that could be very reactive toward the solute, is localized on a relatively unreactive chloride ion. On the other hand, the cationic solvent ions still remain free to exchange charge with solute molecules, forming positively charged solute ions. If the charge transfer is non-dissociative, a radical cation is formed.

For example, the radiolysis of aromatic hydrocarbons in solids^^ , 3 2,3 3 ,34 halogenated hydrocarbon liquids^-^ ’ resulted in the formation of aromatic radical cations through the non-dissociative electron transfer reaction between the solvent radical cation and aromatic solute:

(5) By irradiating precursor compounds such as (PhCH2)2Hg,

Ph^CHBr, and Ph^COH in 1,2-DCE solutions, carbénium ions

such as benzyl cation, benzhydryl cation, and trityl cation

were formed, respectively^^. The trityl cation and benzhydryl

cation were identified by comparison with their reported

optical absorption spectra in sulfuric acid. Since the

optical absorption spectrum of the benzyl cation had not

been previously reported, its identification was based on the scavenging studies and the observation that a single

common absorption spectrum was obtained from a series of

diverse benzyl precursor compounds. The optical absorption

spectra of benzyl cation (with absorption peak at nm),

benzhydryl cation (with absorption peak at 44-9 nm), and

trityl cation (with absorption peaks at 434 and 409 nm)

are summarized in Figure 2.

As an analog to the formation of radical cations in

1,2-DCE, these carbénium ions were thought to be formed by

the dissociative charge transfer reactions between solvent

radical cations and solutes. Taking dibenzylmercury as an

example:

DOE- ■+ (PhCH2)2Hg ---- > DOE + PhCH® + PhCH^Hg. (6)

It will be shown in this dissertation that dissociative

electron transfer reaction is not the only process leading

to the formation of carbénium ion and that the radical cation 22

PhCH® 10

280 320 360 400 480 320 Wavelength, nm

Figure 2. The optical absorption spectra o f (1) benzyl cationC*), resulting from irradiation of a 0.0055 M solution of dibenzylmercury in 1,2-DCE, (2) benzhydryl cation(o), resulting from irradiation of a O.OO96 M solution of benzhydryl bromide in 1,2-DCE, and (3) trityl cation(x), resulting from irradiation of a 0 .II5 M solution of trityl alcohol in 1,2-DCE. 80 ns pulses were used. 11 of 1,2-DCE is not the only species responsible for the formation of carhenium ions.

The absolute reactivity of these carbénium ions toward various nucleophiles such as halide ions, ammonia, amines,. water, and alcohols has been studied previously^^"^^. This work extends the investigation to the study of their reactivity with various alkenes. It is interesting because the reaction between a carbénium ion and an alkene molecule is the initiation step of a cationic polymerization reaction. EXPERIMENTAL

The general method of pulse radiolysis has been discussed

The following descriptions present details of the specific procedures and equipment involved in the

present work.

PULSE IRRADIATION

The source of high energy electrons was a Varian

linear accelerator, Model V-7715-A. The simplified

acceleration process is as follows^. High voltage dc power

is transformed to pulses of radio frequency power by means

of a 3 .0 GHz s-band microwave magnetron rated for 2 megawatts

peak pulse power. The pulses of microwave frequency power

from the magnetron are coupled to the acceleration section, which consists of disc-loaded waveguides. The electrons are

injected in pulse into the travelling microwave field in the waveguide for the period of time selected by the inflector

system. The electron pulse width may be varied over the range from 15 nanoseconds to lAOO nanoseconds. As the electrons enter the microwave field region they are accelerated along

the waveguide and finally pass out of the guide through the quadrupole focus system to the target.

The accelerator can be operated in either the steady

12 13 state mode or the stored energy mode. In the steady state mode, the electron pulse is long compared with the period of time required to propagate the microwave along the full length of the waveguide. The rate of energy extraction hy the electron beam equals the rate of energy input from the magnetron. Thus, a steady state is reached. The pulse duration in this mode ranges between 0 .1 and 1.4- microseconds.

When tuned for maximum efficiency, the mean particle energy is about 3.'5 Mev with 70^ of the electrons being within +0.6 Mev of this value. The maximum current for this mode is about 320 milliamps. The total energy in the beam with a

1 .4 microsecond pulse is about 1 .6 joules per pulse. In the stored energy mode, the electron pulse duration is comparable with the microwave propagation time. The electrons then extract stored energy from the microwave field already existing in the waveguide. The result is a narrow pulse with higher current than the steady state pulse, but with a much broader energy spectrum of the electrons.

The energy spread over the pulse ranges from about 3*^ to 5.6

Mev. The duration of the stored energy pulse is from 15 to

80 nanoseconds with peak current about 600 milliamps. The total energy in a 60 ns pulse is about 0.l6 joules. The higher current and short pulse duration of the electron beam in the stored mode allowed measurable quantities of transients to be formed in shorter times and thus achieved better time resolution than the steady state mode permitted. In this 14 work all the kinetics experiments and some of the spectral measurements were performed using the stored energy mode.

The long term stability of the pulse width in the range

from 100 to l400 nanoseconds is better than 3^ of its setting.

The long term stability of the current setting is about Sfo.

Thus the variation of dose delivered should not exceed 10#

over a four hour period.

The rise and decay time of an electron pulse are each about 6 nanoseconds, independent of pulse width and mode

of operation. Time profiles for two pulses are shown in

Fig 3. Boag^ has discussed the influence of non-uniform distributions of reactants on the measurements of rate cons

tants by the optical absorption method. Under conditions of first order kinetics, variation of transient concentration along the light path has no effect on the rate constant measured; variation of concentration across the light path

causes error less than Ifo in general. Determination of rate

constants under second order conditions is affected by

concentration variation both along and across the analyzing light beam. In most practical cases, the error will be less than 3fo.

OPTICAL DETECTION

The detection of radiation-induced transient species was 15

'T"

---k: J-

Figure 3- Time profile of an 85 nsec pulse (a ) and a 6OO nsec pulse (b) 16

done by fast optical absorption. Schematic diagrams

of two different setups are shown in Figure k . A double pass

optical detection system which was used in most of the kinetic

and spectroscopic studies is shown in Figure 4(a) ; a single-

pass detection system which was used in the temperature

dependent studies is shown in Figure 4(b). In Figure 4(a),

a partially reflecting-partially transmitting mirror was used

in order to monitor two different wavelengths simultaneously.

The mirror was made of quartz and coated with chromium. The

transmitted beam and the reflected beam were directed to two

different detection systems. When it was desired to monitor

one wavelength only, this mirror was removed, along with the

second monochromator and detector. In Figure 4(b), a cryostat

was used to adjust the temperature. The construction of the

cryostat prevented reflecting the analyzing beam through the

cell for a double pass. Therefore, whenever the cryostat

was used, the single pass detection system was employed. The

individual components of the optical detection system are as follows:

The source of analyzing light for all experiments was

an Osram high pressure xenon arc lamp, type XB0450. During

steady state operation the lamp was operated at 20 volts, and

24 amperes current. Output of the lamp could be increased

by discharging large amounts ( 200 A ) of current through it

in a pulse. Variable amounts of charge were stored in a

capacitor bank and were discharged in parallel to the steady 17

Shutter Xenon Mirrors Lamp Detector Lens. To Oscilloscope ^ F ilte rs

Monochromator

FI Lens , Filter cWoLens Monochromator Monochrome / t.j j...I olQ) W Shutter I 1 Det Xenon Cryostat Lamp

Figure ij-. (a) Schematic diagra:.. of the double pass optical detection system using a beam splitting mirror, (b) Schematic diagram of the single pass detection system used in the determination of activation energy. 18 state power supply of the lamp. The maximum resulting increase in intensity was approximately a factor of 10 to a factor of

40, depending on wavelength. The increase was greater toward the U.V.

Spectral radiant energy distribution of the xenon lamp is shown in Figure 5• A sample of the time profile of the lamp flash at 405 nm is shown in Figure 6. At oscilloscope sweep rates of 2 microsecond/division with vertical expansions of 20 mv/division, the 20 microseconds interval centered at the peak of the lamp flash appears flat within an uncertainty of a few percent. This flat portion has to coincide with the time interval of the chemical process being studied. This was done by timing the electron pulse with the start of the flat portion utilizing a 4 millisecond time delay device.

A shutter, which could be operated by remote control, and appropriate Corning filter were placed between the lamp and the reaction cell to minimize photolysis. Filters were also placed before the entrance slit of the monochromator to prevent second order components of light from reaching the photomultiplier.

In many cases, this filter also helped to reduce the Cerenkov emission reaching the detector.

The following Bausch and Lomb monochromators were used:

Type 33-86-01, wavelength region 100-400 nm, 2?00 grooves/mm, dispersion 3 . 2 nm/mm of exit slit width; Type 33-86-02,

350-800 nm, 135O groves/mm, dispersion 6.4 nm/mm; Type 33-86-07,

200-700 nm, 1200 groves/mm, dispersion 7.4 nm/mm; and Type 20

300 400 500 600 700 800 900 1000 1100 wavelength, nm

Figure 5, Spectral radiant energy distribution for Rudolph xenon lamp with 450 watt bulb. I 0 .2 msec

Figure 6. Time profile of xenon flash at 405 nanometers. 21

33-86-0 3 , 700-1600 nm, 6?5 groves/mm, dispersion 1 2 .8 nm/mm.

Calibration of each grating was accomplished by comparison of

the observed wavelengths of known emission lines to their reported wavelengths. Oriel spectral lamps were used as the

sources of the emission lines.

Upon exiting the monochromators, the light was detected by a photomultiplier tube which was shielded by one-half inch of lead to reduce radiation interference. The RCA 7102, HTVI9 6,

7 2 0 0, and 1P28 photomultipliers, with s-1, s-1, s-19, and s-4

spectral responses, respectively, were used. The outputs of

the RCA 7 1 0 2, 7200 photomultipliers were coupled to the oscilloscope through a Nexus Model FSL12 signal conditioning amplifier which allows direct coupled steady state operation into a 93 ohm transmission line. The output of the system was found to be linear with light intensity input in the range of 20 mv to 1 volt. The 10-90^ rise times were approximately

7O-IOO nanoseconds ( essentially the rise time of the Nexus amplifier).

The RCA 1P28 and the HTVI96 photomultipliers were directly

coupled to the 93 ohm transmission line. By eliminating the amplifiers, the 10-90^ rise times of the detection systems using these phototubes were less than 10 nanoseconds. They were linear in output from 0 volt to 1 volt into 93 ohm loads.

The linearity of the output' with light intensity input was checked in the following way. A square wave light pulse with constant low intensity from a LED triggered by a Tektronix 22

114 pulse generator was beamed into the photomultiplier along with a high intensity, continuous light from xenon lamp, and

the amplifier output was observed on an oscilloscope. As the

intensity of the continuous light was varied (so the output"

could be changed from 0 to 1 volt), apparent variation of

the amplitude of the square wave indicated the non-linearity.

The rise time of each detector was measured by observing the Cerenkov light emission. Since the time profile of the

Cerenkov emission follows the electron pulse, its time profile is a rectangular wave with less than a 6 nanoseconds 10-90^ rise time. The observed rise time of the Cerenkov emission thus was taken to be the rise time of the detector systems using

each of the photomultipliers.

The spectral responses of these detectors are shown in

Figure 7. The oscilloscope used in each experiment was a Tektronix,

Type 7844. The bandpass of the scope, including its amplifiers, was 100 MHz. The calibration of the oscilloscope was tested during each experiment. A 0.000-1.000 volt precision voltage source was used to calibrate the gain of the vertical amplifiers and a Tektronix, Type 184, time mark generator was used to calibrate the time bases.

REACTION CELLS

Irradiation cells, from Pyrocell Inc., were made of fused 23 ,to 150

IOC *

300 700 H O C

300 700 1100

300 700 1100 wavelength, nm

Figure 7. Spectral responses of phototubes with (a) s-1 response (b) s-5 response (c) s-1 9 response. 24

silica and had optical windows of high purity fused silica.

The windows were resistant to darkening by radiation. The

internal dimensions of the irradiation cells were 8 mm in the

direction of the electron beam, 20 mm along the analyzing light

beam, and 12 mm in the vertical direction. The quartz walls were approximately 1 mm thick. It has been demonstrated-^

that the total dimension of the electron beam was an oval

shape with approximately 3 cm in width and 2.5 cm in height

as judged by the radiation discoloration of a sheet of

polyvinylchloride. For this type of reaction cell, the

radiation energy deposited is approximately uniform over the

cell area in the plane perpendicular to the electron beam.

■ It has also been determined that the absorbed dose along the

electron beam does not vary by more than 10^ from the value

at the front of the cell"^.

The irradiation cell was just part of the reaction, cell.

Three different reaction cells were used depending upon the

type of experiment in which it was to be used. For spectral

studies, the cell shown in Figure 8(a) was used. The fused

quartz cell was connected to a 50 ml reservoir, and a teflon

stopcock with 0-4 mm bore, from Kontes Glass Co., was used.

It could sustain a vacuum of 10“-^ torr with its two Viton

0-rings. The 9 mm Solv-Seal joint permitted the preparation

of the cell on a vacuum line.

The reaction cell illustrated in Figure 8(b) was used

for rate constant measurement, yield curve study, and 25 9 mm joint

stopcock

cell

pipette

thermoCOupl(

Figure 8. Reaction cells used in this work, (a) spectral cell (h) kinetic cell (c) modified kinetic cell for temperature dependent study. ■ 26 competition kinetic study. Two reservoirs were connected

"by a pipette with teflon stopcocks at each end which permits measured volumes of a solution containing a reactant to be added, through the pipette, to the irradiation section of the cell.

In most experiments the reaction cells were kept at the ambient temperature of the accelerator facility, 23+2°C. When it was necessary to change the temperature as in the case of activation energy determination, a modified kinetic cell shown in Figure 8(c) and a cryostat shown in Figure 9 were used.

The cryostat was a brass box insulated with one inch thick

Styrofoam on its exterior. Two holes, which were aligned on opposite faces of the box, were covered with evacuated quartz cuvettes. The cuvettes served as optical windows for the analyzing beam. The electron window was made of thin copper foil which was situated on one side of the cryostat so that the electron beam would intersect the analyzing beam at a right angle. The kinetic cell was mounted to the cryostat by means of a 34/45 standard taper. An outer taper was an integrated part of the top of the cryostat. A Pyrex inner taper was blown onto a standard kinetic cell as shown in

Figure 8(c). The distance between the irradiation cell and the taper was chosen to ensure proper positioning of the irradiation cell with respect to the electron beam and the analyzing beam.

Cooling and heating was achieved by blowing cold and hot 27

electron beam

light path

cold or hot N,

Figure 9. A cryostat used in temperature dependent experiments, (a) thin copper window, (b) evacuated quartz cell, (c) reaction cell, (d) nitrogen gas outlet. 28 nitrogen into the cryostat. Temperature was regulated by the

rate of flow of nitrogen. It was measured by means of a

copper-constantan thermocouple and a Leeds and Northrup potentiometer. The reference junction was kept at 0°C and the

test junction was cemented to the irradiation cell.

All the reaction cells and storage vessels were cleaned

by the following procedure; The glassware was first rinsed repeatedly with reagent grade acetone. Reagent grade methanol

or absolute ethanol then was used to remove the acetone and

to provide further cleaning action. The glassware was then

placed under vacuum ( 1 0 ~ ^ - 10”^ torr ) for about two days.

SAMPLE PREPARATION

If the solute to be introduced into the cell was a liquid, it was always degassed and stored under its vapor in a glass

storage bulb. The desired amount of liquid measured by

weight difference was then transferred to the reaction cell

by vacuum distillation if the vapor pressure of the liquid

was high enough. In case the liquid had very low vapor

pressure, it was transferred using a pipette and the quantity

transferred was determined by weight difference.

If the solute was a solid, the desired amount was weighted

on an analytical balance. It was transferred into the

reservoir of the cell by pouring through the bore of the

stopcock by means of a funnel. The cell was then joined to

the vacuum manifold and evacuated. The desired amount of

solvent was then vacuum distilled into the cell. 29

Several solutes used in this work including ammonia, ethylene, propylene, isobutylene, and 1,3-butadiene were gases. The preparations of their solutions and calculations of concentrations are as follows;

Since the solubility of ammonia was high in 1,2-DCE

(7 .9 7 mol^ at 20°C under 1 atm^), it was assumed that ammonia was completely dissolved in a reaction vessel in which 80^ or more of the volume was filled with 1,2-DCE. Under such condition, the calculation shows that the ratio of the number of ammonia molecules in the liquid phase to the number in the gas phase is about 100 to 1 by assuming that Henry’s law holds.

■ Ammonia was first transferred from a gas cylinder to a storage bulb of known volume in which the vapor pressure was measured by a calibrated Wallace and Tiernan pressure gauge.

Model FA-160'^. The amount of ammonia was calculated by the ideal gas law.. It was then frozen into the reaction cell in which at least about 80% of the volume was occupied by 1,2-DCE.

The solubilities of the above mentioned alkenes in

1,2-DCE were not known. Their solubilities were measured by the following method. A high vacuum manifold equipped with pressure gauge and gas inlets as shown in Figure 10 was used for all the measurements. First a storage bulb of known volume was filled with alkene gas, the pressure of which was measured by the pressure gauge. This known amount of 30

high rough

pressur( Teflon stopcock gauge

inlet |Storage bulb

1,2-DCE

Figure 10. Apparatus used for the determination of solubility of gas in 1 ,2-dichloroethane solution. 31 gas was transferred to another known volime storage bulh A containning known volime of 1,2-DCE. The storage bulh A was then joined to the high vacuum manifold at the position of stopcock 0 as shown in Figure 8. After evacuating the shaded part of the vacuum manifold, stopcock a and b were closed and stopcock c was opened, vapor pressure was measured again after the whole system reached equilibrium. By subtracting the vapor pressure of pure 1,2-DCE, from this total pressure, P^, the pressure of alkene was obtained. The amount of alkene gas in the gas phase was then calculated according to the following equation:

^ _(Pt ~^dc e ^(Vs+V^-Vd c e ^

where = volume of shaded part in Figure 8.

= volume of storage bulb A.

Vd c e “ volume of 1,2-DCE in bulb A. R = Boltzmann constant.

T = temperature.

Since the total amount of alkene gas in the system was known, the amount of alkene dissolved in 1,2-DCE was obtained.

In the above equation, the sum of was obtained as follows: The empty bulb A was joined to the vacuum manifold at position c as shown in Figure 8 and evacuated. With stopcock a closed, b and c opened, alkene was traAferred into the vacuum line and its pressure measured. After closing 32 stopcock c and evacuating the shaded part, stopcock a and b were closed and stopcock c opened again. New vapor pressure was measured after the system reached equilibrium. The sum of Vg and was then obtained from the equation: PV=constant at fixed temperature.

According to the above procedure, the solubilities of ethylene, isobutylene, and 1 ,3-butadiene in 1,2-DCE were measured as shown in Table 1.

Table 1. Solubilities of alkene gases in

1,2-DCE at 25°C.

solubility

ethylene 0 .9 5 mol^ at 9 2 7 .8 torr isobutene 1 .0 7 mol^ at 131 torr

1 ,3-butadiene 1.08 mol^ at 9 0 .2 torr

For the experiments with isobutene, 1,3-butadiene, and propylene, 90fo or more of the reaction cells were occupied by 1,2-DCE. For the experiments with ethylene, at least

$8% of the reaction cell was occupied by 1,2-DCE. Under these conditions, the calculation shows that at least 99f° of the alkene gas in the reaction cell was dissolved in 1,2-DCE.

The procedures to transfer the alkene gases into reaction cells were the same as described under ammonia. 33

DOSIMETRY

The dosimetry method described by Dorfman and Taub® was used. The solution was the modified Fricke dosimeter contain ing 10“^ M ferrous sulfate and 0.4 M sulfuric acid. The solution was saturated with oxygen and contained no chloride ion. Dose was calculated using 220 M“^cm“^ as the extinction coefficient at 25°C for the ferric ion at 304 nm and 15.6 molecules/lOO ev as the G value of ferric ion.

DATA ANALYSIS

Data from the pulse radiolysis experiments were always obtained in the form of Polaroid photographs of oscilloscope traces. Either Type 107 film, 3000 speed, or Type 4l0 film,

10,000 speed, was used, depending on the intensity of the trace that could be obtained.

Figure 11 shows two representative oscilloscope traces.

Figure 11a represents the formation process of a transient species, and Figure lib represents a decay process.

The portion of the trace proceeding the electron pulse represents the amount of light, I^, transmitted through the unirradiated cell. This value was measured by a sampling digital volt meter, a " I^ meter ". This " I^ meter " could sample the amplitude of a signal over a 1 microsecond interval. 34

,electron pulse

200 nsec

.electron pulse

I, 1 psec

Figure 11. Two typical oscilloscope traces showing formation (a) following the pulse and decay (b) following the pulse. 35

The end of the 1 microsecond sampling interval was placed 200 nanosecond before the start of the electron pulse for all

experiments reported herein. The meter was capable of reading within 1 mv for a typical 800 mv signal.

The relative intensity of the light that was absorbed at

any time, 1 ^ , was measured by a Hewlett Packard digitizer,

Model 9107A. The digitizer consists of a cursor, a platen, and a mainframe. The cursor is an electro-magnetic transmitter,

operating at 3 KHz. The platen, a digitizing surface, acts as

a receiving system. Conversion of the received information

into coordinates describing the position of the cursor occurs

in the mainframe. The kinetic curves shown in Figure 9 were

digitized by tracing over them with the hand held cursor.

Data storage and program execution were done by a Hewlett

Packard calculator. Model 9100B and a Hewlett Packard Model

9101A extended memory. Kinetic plots and the resulting rate

constants were obtained directly from a Hewlett Packard 9123A plotter and a 9120A printer.

The quantities, and 1 ^ , that were measured from

oscilloscope pictures were related to concentration by the

Beer's law.

I . A Â TÂ- = E |c]l where OD = optical density at wavelength A . 36

1 = optical light path.

[cj= concentration of absorbing species.

Because of the transient nature of the absorbers studied in this work, the validity of Beer's law was not directly tested. However, the applicability of this law to a similar pulse irradiation system has been carefully examined.^

It was shown that the law is valid providing £ is nearly constant over the bandpass of the monochromator. Since the bandwidth from the monochromators was typically less than

3 nm, which was small compared to the half-width of the absorption spectra reported in this work (about 60-100 nm),

Beer's law was thus assumed valid.

1 ,2-Dichloroethane was obtained either from Matheson,

Coleman, and Bell, reagent grade or from Aldrich, 99+^» spectrophotometric grade, gold label. The 1,2-DCE from

MOB was purified in the following manner; Approximately

80 ml of concentrated sulfuric acid were added to 400 ml of 1,2-DCE and the mixture was vigorously shaken in a separatory funnel in order to remove alcohol added as a stabilizing agent. The acid was removed and the 1,2-DCE washed at least three times with demineralized double distilled water. It was then washed twice with saturated 37

aqueous solutions of sodium carbonate. Following four or five

washings with double distilled water, the 1,2-DCE was then

washed with triply distilled water several times. It was

refluxed over for about two hours and then distilled into

a storage bulb; the first 1^% and the last 15^ were discarded.

After repeated freezing-evacuating-thawing cycles, it was then

vacuum distilled into another storage bulb containing freshly

prepared Linde type k k molecular sieve where it was stored

under vacuum and dark. The molecular sieve was prepared by

heating at about 200-300°C under vacuum (10“^ torr) for about

two days. The 1,2-DCE obtained from Aldrich was stored under nitrogen and free of any alcohol. Therefore the washing

steps with acid, base, and water were omitted during purifica

tion. No difference was observed between experiments using

1,2-DCE from these two different sources.

Bromodiphenylmethane, technical grade from Chemical

Samples Co., was purified by vacuum sublimation twice. The

pure sample was white, and the crude sample was brown.

Benzhydrol (995^. from Aldrich Chemical Co.) was purified

by recrystallization from a solution containing equal volumes of triply distilled water and absolute ethanol (U.S.?., from

Commercial Solvents Co.). Hot solution was poured onto the

sample until it was completely dissolved. The solution was filtered and the filtrate allowed to cool. The pure crystals

that appeared upon cooling were separated and dried by filtra

tion in a Buchner funnel. Further drying was accomplished by 38 pumping on a vacuum line.

Triphenylcarbinol (97?^, from Aldrich Chemical Co.) was recrystallized from either 1,2-DCE or cyclohexane (gold label, spectrophotometric grade, from Aldrich). The procedure was the same as that for benzhydrol with the additional step of adding activated charcoal to the hot solution of dissolved sample.

Triphenylmethyl bromide (J.T.Baker Chemical Co.) was purified by recrystallization from 1,2-DCE according to the procedure outlined for benzhydrol.

Dibenzyl mercury ( from Alfa Inorganics ) was purified by recrystallization from absolute ethanol (U.S.P. from

Commercial Solvents Co.). The purification outlined for benzhydrol was followed.

Anthracene, 99-9+^, gold label, was obtained from

Aldrich and was used without further purification.

All the above compounds were stored under vacuum and in the dark.

The following gases were obtained from Matheson Gas

Products and were used without further purification:

Ammonia- 99-99^ minimum, anhydrous grade. Lecture Bottle;

ethylene- C.P. grade, 9 9 .5 % minimum, Lecture Bottle; 1,3-butadiene- instrument purity, 9 9.5% minimum. Lecture Bottle; propylene- C.P. grade, 99-0^ minimum. Lecture Bottle.

Isobutene, research grade, 99«93^, was from Phillips

Petroleum Company and was used as obtained. 39

The following alkene compounds were obtained from

Aldrich Chemical Co.: 3,3-Dimethyl-l-butene ( neohexene ),

99^; cyclohexene, 99^; 1,3-cyclohexadiene, 99^; 2-methyl-

1 .3-butadiene ( isoprene ), 99+^. gold label; 2,3-dimethyl-

1.3-butadiene (DMBD), 98^; 4-methyl-l ,3-PGntadiene (BdPD), 98^;

2.4-dimethyl-l,3-pentadiene (DRIPD), 9 8 fo . All of these

chemicals were purified by vacuum distillation and stored under their own vapor in a refrigerator.

1,1-Diphenylethylene., 97^ and allylbenzene were obtained

from Aldrich and were purified by partial freezing. They were also stored under their vapor in a refrigerator. RESULTS AND DISCUSSIONS

I. MECHANISM OF THE FORMATION OF ARILCARBENIUM IONS IN

THE IRRADIATED 1.2-DICHLOROETHANE SOLUTION

It has been well established that molecular radical cations of aromatic compounds (e.g. biphenyl, anthracene, and p-terphenyl) can be efficiently produced by the radiolysis of their alkyl halide solutions in both solid 31-3^»37,38 liquid^^’^^ phase. It is generally recognized that the first step of radiolysis is the formation of solvent radical cations.

The solvent radical cations then exchange their positive charges with aromatic solutes through non-dissociative charge transfer process.

+ Am » S + Am^ (7) where st represents solvent radical cation and Am represents aromatic solute. The mechanism of the positive charge migration has recently been examined by Arai, Kira and Imamura in alkyl halide matrix at low temperature^^,3^,39 ^ Their results will be brought into discussion later in this section.

The formation of arylcarbénium ions by the radiolysis of appropriate solutes in 1,2-DCE solutions has also been demonstrated^^. As an analog to the formation of molecular radical cations, it was thought that they were formed by 40 4l dissociative charge transfer reaction. Taking henzhydryl bromide as an example:

Ph^CH® + Br- + S (8)

The mechanism of the formation of arylcarbenium ions

in irradiated 1,2-DCE solutions at 24°C has now been examined

in detail. The experimental evidence that will be presented in this section indicates that while reaction (8) ocours, it is not the only process responsible for the formation of arylcarbenium ions in irradiated 1,2-DCE solutions.

Three different approaches were used in this study.

They will be discussed in the following.

Yield curve studies

The first step of the radiolysis of 1,2-DCE is the

formation of solvent cations, S"*". In the absence of any solute, the solvent cations decay via charge recombination processes

with chloride ions and/or other processes involving solvent

itself or even impurities. With the addition of appropriate solute, the solvent cations react with solute and form

carbénium ion.

+ X ---> product (9)

+ P ---> P® (10) 42 where S'*’ is a general term for any cationic species derived from solvent. Its use here is intended to be ambiguous and will be clarified later. P and P® represent precursor compound and carbénium ion respectively. X represents chloride ion, solvent molecule, or impurities.

According to the above scheme, when the concentration of solute increases the yield of carbénium ion also increases as a result of the competition between reaction (10) and reaction

(9). The yield of carbénium.ion will finally reach a plateau as the concentration of solute becomes so high that all the solvent cations are scavenged to carbénium ions (i.e. reaction

(10) becomes the dominant process). Therefore, the concentration of carbénium ion at the plateau of the yield curve is equal to the concentration of solvent cation. If a single solvent cation is responsible for the formation of two carbénium ions from two different precursor compounds, then the plateaus of the two yield curves should have the same height (i.e. correspond to the same concentration).

The yields of benzyl, benzhydryl,. and trityl cation as functions of the concentrations of their precursor compounds have been studied. The results are shown in Figure 11, l6, 17»

20, 22, and 23. In each experiment the optical density of benzyl cation was used as a standard to monitor the dose variation. The relative yields of arylcarbenium ions were then calibrated to the same dose level. This ensures that the yields of different carbénium ions may be compared without the 43 consideration of dose variation in each experiment. Any complication due to overlapping absorption has also been taken into consideration in order to obtain a correct value.

Different methods were used to correct for overlapping absorption depending upon the precursor compounds. In the following the procedures to obtain the results will be presented and then the significance of the results will be discussed.

(1) Benzyl cation from dibenzylmercury.

Irradiation of dibenzylmercury-DCE solution generates benzyl cation which has an absorption peak at 364 nm. Figure 12 shows the plot of optical density taken immediately after the pulse at 364 nm versus the concentration of dibenzylmercury.'

Most of the absorption at 364 nm wavelength is due to the benzyl cation. Nevertheless, there is still significant amount of overlapping absorption from a species with an absorption peak at 323 nm as shown in Figure 13. This unknown species has a longer lifetime (about 20 microsecond) than benzyl cation

(about 7 microsecond). It is not scavenged by oxygen and alkene, but can be scavenged by ammonia. To correct for the overlapping absorption, enough concentration of alkene such as propylene was added to remove the absorption due to benzyl

cation. Time dependent spectrum of irradiated propylene- dibenzylmercury-DCE solution as shown in Figure 14 indicates

the absorption at 364 nm is mainly due to the unknown species 9 3 8

7 6

0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 [(PhCHp)2Hg]x 10^,

Figure 12 . The plot of optical density at 364 nm vs. the concentration of dibenzylmercury in the irradiated dihenzylmercury-1,2-DCE solution. A 80 ns pulse was used. The points were taken immediately after the electron pulse. The dashed line indicates the plateau. PhCH, 10

t=end of pulse g

280 290 300 310 320 330 340 350 360 370 380 390 400 410 420 430 440 450 Wavelength, nm Figure I3 . The time dependent spectrum resulting from irradiation of a 5.5xlO“^M solution of dibenzylmercury in 1,2-DCE under vacuum. A 80 ns pulse was used. 9

8 PhCH 7

3 t=end of pulse

0 310 320 330 340 350 360 370 380 390 400 410 420 430 440 450 460 Wavelength, nm

Figure 14. The time dependent spectrum of irradiated propylene-dihenzylmercurys 1,2-DCE solution under vacuum. [ propylene]=0.742 M. (dibenzylmercury)=4.9x10"%. A 80 rts pulse was used. 47 after about 1.4 microsecond. The decay curve of benzyl cation in the presence of 0.742 M propylene is also shown in Figure

15 . It is apparent that benzyl cation disappears within about

1.4 microsecond. By using propylene and isobutylene as scavengers, the contribution of the unknown species to the O.D. at 364 nm at the plateau was determined to be 0.018 for an

80 ns pulse. The true O.D. of benzyl cation at the plateau was then obtained by substracting 0 .0 1 8 from the total optical density at 364 nm.

In order to calculate the concentration of benzyl cation from its optical density, the extinction coefficient has to be known. Unfortunately, the extinction coefficient of benzyl cation has not been reported in the literature. The following method was used to determine its value. Since the natural decay of benzyl cation is governed by the charge recombination reaction with chloride ion. The plot of l/OD of benzyl cation versus time should give a straight line with

slope = (11) 1-6 where OD represents the optical density, k^ is the second order rate constant between benzyl cation and chloride ion,

1 is the optical path, and £ is the extinction coefficient. An example of the plot is shown in Figure 16. The average slope from several of these plots was determined to be 1 .55x10^ sec“^.

Although the rate constant of the reaction between benzyl cation 48

768 mv

50 mv

500 ns

Figure I5. The decay curve of "benzyl cation in the presence of 0 .7 4 2 M propylene. The wavelength monitored was 369 nm. A 80 ns pulse was used. It is apparent that some residual absorption exists. 49

Time

Figure 16. The plot of l/optical density of benzyl cation at 364 nm as a function of time in the irradiated dibenzyl mercury- 1 ,2-DCE solution. A 80 ns pulse was used. 50

between benzhydryl cation and chloride ion, 9*0 x 10^^ M~^sec~^,

and between trityl cation and chloride ion, 8 .0 x 10^^ M"^sec"^. All of these reactions are believed to be diffusion-controlled.

The extinction coefficient of benzyl cation was thus

calculated to be 16OOO ± M~^cm~^. By using the corrected

optical density and calculated extinction coefficient, the

concentration of benzyl cation at the plateau of the yield

curve was determined to be 1.03 x 10”^M +20^ for an 80 ns. pulse. This is shown in Figure 23.

(2) Benzhydryl cation from benzhydryl bromide.

Irradiation of benzhydryl bromide-1,2-DCE solution forms

benzhydryl cation with absorption peak at 449 nm. Figure 1?

shows the plot of optical density taken immediately after the pulse at 449 nm as a function of the concentration of benzhydryl bromide. A plateau is clearly defined for concen

tration higher than 7 x 10"%. The absorption at 449 nm is due only to benzhydryl cation. By using the reported^® extinction coefficient of benzhydryl cation, 38000 M“^cm“^,

correcting for the dose variation as described earlier, the

concentration of benzhydryl cation at the plateau of the

yield curve was determined to be I .30 x 10“% for an 80 ns,pulse. This is shown in Figure 2 3. 51

5 6 7 8 9 10 11 12 13 14 [PhgCHBr] X 10^, M

Figure 17. The plot of optical density taken immediately after the pulse at 449 nm as a function of the concentration of benzhydryl bromide. A 80 ns pulse v/as used. The dashed line indicates the plateau. 52

(3) Benzhydryl cation from benzhydryl alcohol.

Irradiation of diphenylmethanol-1,2-DCE solution forms benzhydryl cation. Figure l8 shows the yield curve with points taken immediately after the pulse. A plateau is reached for concentration higher than 2 x 10”^ . Figure 19 shows the rate curve of benzhydryl cation at 449 nm. A residual absorption with optical density about 0.007 exists at the tail of the rate curve. This residual absorption has to be sub- stracted from the total optical density at 449 nm in order to obtain the true optical density of benzhydryl cation at this wavelength. By using the same reported^® extinction coefficient,

38000 M”^cm“^, the concentration of benzhydryl cation at the plateau of the yield curve was calculated to be 0.34 x 10“^ M

+ 15fo for an 80 ns pulse. This is shown in Figure 2 3.

(4) trityl cation from trityl alcohol.

Irradiation of trityl alcohol-1,2-DCE solution forms trityl cation with one of the absorption peaks at 434 nm. Figure 20 shows the plot of optical density taken at 1 microsecond after the pulse at 434 nm versus concentration of trityl alcohol. Figure 21(a) shows the rate curve of trityl cation at 434 nm. It is clearly that there is a fast decay in the beginning of the rate curve. This fast decay is due to the overlapping absorption from some species which absorbs at longer 10 11 12 13 14 15 16 [phpCHOH]x 10^

Figure 18. The plot of optical density at 449 nm as a function of the concen tration of diphenylmethanol in the irradiated diphenylmethanol-1,2-DCE solution. A 80 ns pulse was used. The points were taken immediately after the electron pulse. The dashed line indicates the plateau. 54

50 mv

Figure 19. The rate curve of benzhydryl cation in the irradiated diphenylmethanol -1,2-DCE solution. The wavelength monitored was 449 nm. A 80 ns pulse was used. The existence of residual absorption is obvious. 55

11 . 12 13 14 15 [Ph^COH] X 1 0 ^ , M

Figure 20. The plot of optical density at 434 nm as a function of the concentration of triphenylmethanol in the irradiated triphenylmethanol-1,2-DCE solution. A 80 ns pulse was used. The points were taken 1 ps after the pulse. The dashed line indicates the plateau. 56

830 mv

50 mv

1 n s

(b) 40 mv

Figure 21. (a) The rate curve of trityl cation at 434 nm in the irradiated triphenylmethanol-1,2-DCE solution with [triphenylmethanol] = 0.144 M. A 80 ns pulse was used. The initial fast decay is due to the overlapping absorption from (b). (b) The rate curve of the irradiated triphenylmethanol- 1,2-DCE solution at 506 nm. [triphenylmethanol] = O.II5 M. A 80 ns pulse was used. 57 wavelength. Figure 21(b) shows the rate curve of this unknown species at 506 nm. From both Figure 21(a) and (b), it is apparent that this unknown species disappears within 1 micro second while trityl cation does not decay significantly.

Therefore, the optical density taken at 1 microsecond after the pulse is approximately equal to the initial optical density of the trityl cation. By using a value of 40700

M ”^cm“^ as the extinction coefficient of trityl cation, which is about the average of two different reported value, 39810 and 4l68? M“^cm“^ the concentration of trityl cation at the plateau of the yield curve was determined to be 0 .3 2 x 10 -6 M for an 80 ns pulse. This is shown in Figure 2 3.

(5) Trityl cation from trityl bromide.

Irradiation of trityl bromide-1,2-DCE solution forms . trityl cation. The yield curve is shown in Figure 22 which is taken from reference 42. By using 40?00 M"^cm"^ as the extinction coefficient, the concentration of trityl cation at the plateau of the yield curve was determined to be 1 .2 9 x

10"+20^ for an 80 ns pulse. This is shown in Figure 23.

In Figure 23 all of the results obtained above are summarized together. The concentrations of arylcarbenium ions at the plateau of their yield curves are shown to be dependent upon the precursor compound that was used. As mentioned 0.24

0.21

0.16

0.08

Ph^CBr

Figure 22. The plot of the optical density of trityl cation at 439 nm as a function of the concentration of triphenylmethyl bromide in 1 ,2- DCE. A 80 ns pulse was used.• The dashed line indicates the plateau. This Figure is taken from reference 42. 59

Concentration, M x 10 14 12 10 8 6 4 2 0

PhCH:

s 4 6 8 10 12 Concentration, M x 10^

Figure 23. A summary of the yield curve studies. The concen tration of arylcarbenium ion at the plateau of the yield curve is plotted against its precursor compound. A 80 ns pulse was used for all the experiments. All the data were calibrated to the same dose level. 60 earlier, the concentration of carbénium ion at the plateau of the yield curve equals the concentration of solvent cation that can be scavenged by the precursor compound. It is obvious that different solvent cations are involved in the formation processes of arylcarbenium ions and different precursor compounds can selectively scavenge these solvent cations. A closer look at Figure 23 reveals that the yield of benzhydryl cation from benzhydryl bromide is approximately equal to the yield of trityl cation from trityl bromide, the yield of benzhydryl cation from benzhydryl alcohol is approximately equal to the yield of trityl cation from trityl alcohol, and the yield of arylcarb enium ion from bromide compound is approximately equal to the sum of the yield of arylcarbeniurn ion from alcohol compound and the yield of benzyl cation from dibenzylmercury. This implies that there are two types of solvent cations responsible for the formation of arylcarbenium ions. Dibenzylmercury scavenges type 1,

S^, alcohol compounds scavenge type 2 , S^» and bromide com pounds scavenge both.

{ . S g , (12) + (PhCH2)^Hg ---> PhCH® (13)

Ph_COH ---» Ph.,C® • (14) s ; + 3 PhgCHOHPh2C -- > PhgCH"^ (15)

Ph^CBr ---> Ph^c"^ (l6) s;, s: (17) 61

Type 1 ions are likely to be radical cations derived from 1,2-DCE which can perform electron transfer reactions with

solutes of lower ionization potential and are responsible for

the formation of aromatic molecular radical cations in the

solids and liquids. Type 2 species should be some cationic

species different in nature from the radical cation, and the

only reasonable choice seems to be the chloroethyl cation • resulting from the cleavage of C-Cl bond in a 1 ,2-DCE molecule.

More experimental evidences will be presented to support

this viewpoint in the following.

Competition kinetics.

The nature of the cationic species derived from 1,2-DCE

that are responsible for the formation of arylcarbénium ions was also studied by adding another solute to compete with I the formation process of arylcarbenium ion.

scheme is written as follows.

^ + P P® (18)

+ R — product (19)

where S represents solvent cation, P represents precursor

compound, P® represents carbénium ion, and R represents the

second solute added to the system. The concentration of

precursor compound has to be high enough to compete 62 overwhemingly with the natural decay of solvent cation, reaction (9). Usually the concentration of precursor compound was chosen at or near the plateau of its yield curve.

The kinetic equations of the above reaction scheme were solved as shown in Appendix A. The solution is

k„[R]

where OD ^ = optical density of carbénium ion taken at time = P in the absence of the competition solute, R.

OD*0 = optical density of carbénium ion at time = 0 P in the presence of R. [p] = the concentration of precursor compound. [ r J = the concentration of competition solute.

By plotting (OD ^ 0 D*q )-1 versus [Rj/fP] , a straight' line should be obtained with slope equal to Since the formation rate constants of carbénium ions, k^g's, are known^^'^^, the value of k^^ can thus be determined.

If there is only one solvent cation responsible for the

same for different precursor compound and different aryl carbenium ion. The experimental data that will be presented later shows this is not true. The optical density in equation (20) is defined as the optical density at time = 0. Since the value taken immediately 63 after the pulse sometimes is not a good approximation of the value at time = 0 , all of the optical densities were obtained by extrapolation of rate curve back to time = 0. The details of the extrapolation procedure will be explained in each individual case.

In this work, ammonia was used as the competition solute. It is known as a scavenger of cationic species.

(l) Competition between ammonia and dibenzylmercury.

Irradiation of dibenzylmercury-1,2-DCE solution forms benzyl cation with an absorption peak at 364 nm. Addition of ammonia decreases the optical density of benzyl cation at

364 nm as a result of the competition for solvent cation.

The reaction scheme is written as follows.

sj + (PhCHg^gHg PhCH®" (21)

product (2 2)

Figure 24 shows the rate curves of benzyl cation with

[nH^] - 4.4 X 10" % and 7*2 x 10"%. It is clear from Figure

24(a) that at lower concentration of ammonia the optical density at time = 0 can be obtained by direct extrapolation of decay curve back to the time = 0 position. In fact, the optical density taken immediately after the pulse is approximately equal to the optical density at time = 0 in this 64

t=0

(a) 50 mv ,end of pulse

500 ns

t=0 737 mv

(b) of pulse %

100 ns

Figure 24. The decay curves of benzyl cation in the irradiated ammonia-dibenzylmercury-DCE solution. The wavelength monitored was 364 nm. A 80 ns pulse was used, (a) Concentration of ammonia = 4.4 x 10 exp(-5) Concentration of ammonia = 7.2 x 10 exp(-4) 65 case. However, for higher concentration of ammonia. Figure

24(h), the decay rate of henzyl cation is so fast that the portion of benzyl cation consumed during the electron pulse cannot be neglected. Since the decay curve of benzyl cation is a pseudo-first order process representing the following reaction;

» product (23)

the plot of In OD of benzyl cation versus time should give a straight line with slope = and intercept = lnOD^_g.

The OD's at time = 0 were thus obtained from the intercepts of such types of plots. An example is given in Figure 2 5.

Figure 26 shows the plot of ( / 00*^^^®) -1 versus

[nH^]/[(PhCH2)2Hg]. A straight line was obtained with slope

=0 .2 6 5. Since the formation rate constant of benzyl cation from dibenzylmercury, k^^, has been determined^^’^^ to be

1 .3 X 10^^ M"^sec"^, the rate constant for the reaction between ammonia and solvent cation S^, k2 2» "''^as determined to be

3.4 X 10^ M"^sec"^' + 25^.

(2) Competition between ammonia and trityl alcohol.

Irradiation of trityl alcohol-1,2-DCE solution forms trityl cation with absorption peaks at 434 and 409 nm. Addition of ammonia reduces the optical density of trityl cation. The 66

Time

Figure 2 5. The plot of -ln(optical density) of "benzyl cation at 364 nm versus time in the irradiated ammonia-di"benzyl- mercury-DCE solution. [NH«] = 7.2 x 10 exp(-4)M, slope=3.84 X 10 exp(6), intercept = -^2.4l. 8

® CVJ 7

1 0

(NH^]/[(PhCHg)^Hg] X 10

Figure 26. The plot of (0D/0D*)-1 versus ]/[(PhCHp)pHg]. The optical density of henzyl cation was taken at 364 nm. A 80 ns pdlse was used. 68 reaction scheme is

Sg + Ph^COH Ph^C® (24)

+ NK^ — ----- > product (25)

Figure 2? shows the decay curve of trityl cation at 434 nm with [nH^J =5.1 x 10“\ . The initial fast decay is due to the overlapping absorption from the species absorbs at longer wavelength. Because of the lower reactivity of trityl cation toward ammonia, 2.4 x lO*^ M“^sec”^, its decay curve is rela tively flat compared to the fast decay as shown in Figure 2?.

The optical density at time = 0 was thus obtained by direct extrapolation of rate curve to the time = 0 position.

Figure 28 shows the plot of (ODp^ ^,© /ODp^ g@)-l versus

/^Ph^COHj . The points were taken at two different wavelengths, 434 and 409 nm. A straight line was obtained • with slope = 2 7.5. Since the formation rate constant of trityl cation from trityl alcohol has been determined^^’^^

and solvent cation S^, , was thus determined to be 1.6 x

10^° M"^sec"^ ±20^.

At this point it is fairly clear that the solvent cation

S^, which is responsible for the formation of benzyl cation from dibenzylmercury, has different kinetic behavior toward ammonia from the solvent cation Sp, which is responsible for the formation of trityl cation from trityl alcohol. These 69

50 mv

Figure 2?. The decay curve of trityl cation at 434 nm. with[NH«]= 5.1 X 10 exp(-4) M. A 80 ns pulse was used. The initial fast decay is due to the overlapping absorption of some other species. The "o" indicates the position where the optical density at t=0 was taken. 70

20 %

% 14

§ 10

0 1 2 3 4 5 6 ? 8 9 10 11 [NH^]/[Ph^COH] X 100

Figure 28. The plot of (GD/OD )-l versus [NH«]/[PHoCOH]. The optical density of trityl cation was taken at ^ 434 nm (•) and 409 nm (0). A 80 ns pulse was used. 71

results are parallel to the results of yield curve studies.

(3) Competition between ammonia and benzhydryl bromide.

Irradiation of benzhydryl bromide-1,2-DCE solution forms

benzhydryl cation which absorbs at # 9 nm. Addition of

ammonia reduces the optical density of benzhydryl cation at this wavelength as a result of the competition for solvent

cation.

Figure 29 shows the decay curves of benzhydryl cation

with [nH^] = 6.5 X 10“-^M and 1.65 x 10"%. At lower concentra tion, the optical density at time=0 was obtained from direct

extrapolation of the rate curves as shown in Figure 29(a). At higher ammonia concentration, the optical density at t=0

was obtained from the intercept of the plot of In ODp^ versus time.

If the simple reaction scheme, reactions (18) and (19),

is assumed to be valid, then the plot of (ODp^ CH®/*^^Ph CH®^ -1 versus / [Ph^CHBr] should yield a straight line.

Figure 30 shows such plots from two different sets

of experiments. It is obvious that a curve was obtained

instead of a straight line. This is not surprising judged

by the previous results. The yield curve studies suggest

that there are two solvent cations responsible for the forma

tion of benzhydryl cation from benzhydryl bromide. The

competition kinetics studies suggest that these two solvent 72

812 mv __S

(a)

50 mv

^end of pulse

500 ns

t=0 821 mv

(b) 50 mv ^end of pulse

100 ns

Figure 29. The decay curves of "benzhydryl cation in the irradiated ammonia-benzhydryl bromide-DGE solution. The wavelength monitored was 44-9 nm. A 80 ns pulse was used, (a) Concentration of ammonia = 6.5 x 10 exp(-5) M. (h) Concentration of ammonia = 1 .65 x 10 exp(-3) M. 9 8

1 ■fes 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 [NH^J/CPhgCHBrJx 10

Figure 30. The plot of (OD/OD )-1 versus [NHol/lPhpCHBr]. The optical density of benzhydryl cation was taken at 449 nm. Solid dircleS and open circles are from two different experimental runs. A 80 ns pulse was used. 74 cations have differentreactivities toward ammonia. Addition of ammonia in the irradiated benzhydryl bromide-1,2-DCE solu tion will scavenge solvent cation with a rate constant of

1.6 X 10^® M~^sec“^ and solvent cation with a slower rate,

3.4 X 10^ M~^sec“^ . It is this difference in reactivity which causes the curvature in Figure 30.

The reaction scheme should be represented as follows:

PhgCH® (26)

PhpCHBr — ^ PhpCH® (2?) kpm NHo ^product (28) k NH„ -----— > product (29)

The kinetic equations for this reaction scheme can be solved as shown in Appendix B. The solution is

OD* m ^26^^lio ^ ^2?[^2]o °°Ph^CH® k,. + k^pX

where ODp^ = optical density of Ph^CB^ at 449 nm in the

presence of ammonia.

ODph 0^0 = optical density of Ph^CH® at 449 nm in the absence of ammonia.

• t^llo ~ concentration of solvent cation at time=0. 75 concentration of solvent cation at time=0.

X = [NH^]/[Ph^CHBr] .

The yield ciirve studies have determined that ]o/[^2 L = 3«1. The formation rate constant of benzhydryl cation from benzhydryl bromide, , has also been determined^^’^^ to be 1.6 X 10^*^ M“^sec“^. Therefore, equation (30) can be further simplified to equation (31).

3.1(1.6-k2y) (1.6-kgy) + kggX (31) 4.1

In the above equation, k^g and k^^ have been determined to be

3.4 X 10^ M“^sec“^ and 1.6 x 10^*^ M~^sec“^ respectively from earlier results of competition kinetics. In order to plot

/°^PhgCH® "^^^sus [nH^] /[Ph^CHBr] , k^y is the only unknown quantity. The value of k^y can then be obtained by changing its value until a best fit is obtained between computer simulated curve and experimental points. Unfortunate ly, the best fit was not obtained in the region of high value of X, which is mostly controlled by the value of k^g. By changing both the values of k^y and k^g, a best fit was then obtained with k^y = 2.0 x 10^ M“^sec“^ and k^g = 2.5 x 10^

M“^sec“^ as shown in Figure 31. The value of k^g obtained from two different experiments, 2.5 x 10^+i^+ 259g and 3.4 X 10^

M ^sec~^, are within the limit of 10

0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 1^ 15 16 17 18 19 20 [NH^l/lPhgCHBr] x 10

Figure 31. The plot of OD /OD versus (NHo l/[PhpCHBr]. The solid circles are experimentally determined points. The curve was calculated as described in the text. The wavelength monitored was 449 nm. 77 experimental uncertainty. The value of the rate constant

obtained by this method, 1.4 x 10^® M“^sec“^, is the typical value for a diffusion controlled electron transfer 42 reaction.

In summary, the idea that the radiolysis of 1,2-DCE forms two kinetically distinguishable solvent cations which can then be selectively scavenged by different precursor compounds to form different arylcarbenium ions was firmly established by the above yield curve and competition kinetics studies. The natures of the solvent cations remain to be identified. In solution phase this can be done only with the help of gas phase and solid phase data and by some indirect methods that will be described later. From solid phase data^^"^^'^^'^^, it is certain that the radiolysis of 1,2-DCE molecules forms solvent radical cations which are responsible for the formation of aromatic radical cations in both solid and liquid phase.

The solvent cation in this work may be identified with this radical cation of 1,2-DCE. The other solvent cation is probably not a radical cation in nature because of its different kinetic behavior from S^. It may be a chloroethyl cation resulting from the cleavage of C-Cl bond in a

1,2-DCE molecule. This chloroethyl cation may be a carbénium ion or a usually more stable chloronium ion^^“^^.

CHp-CH. (32) \2y 2 Cl 78 47 Recently Beauchamp and coworkers have studied the ionization and fragmentation of 1,2-DCE molecule in gas phase by photoionization mass spectrometry and ion cyclotron resonance spectroscopy. The ions formed are the parent radical cation, ClCH^CH^Cl*, the radical cation CICHCH^», and the cyclic chloronium ion^"^i This is consistent with the con clusion made from the above competition kinetics and yield

would be indistinguishable in this work if they have similar rate constants for electron transfer reactions.

In order to firmly establish that solvent cation is a radical cation in nature and is responsible for the forma tion of benzyl cation from dibenzylmercury, the following competition kinetics experiment between ammonia and anthracene was performed. The existence of chloroethyl cation in irradiated 1,2-DCE solution was also demonstrated by

"titration experiments" that will be discussed later.

(4) Competition between ammonia and anthracene.

Irradiation of anthracene -CCl^^ matrix at 77°K^^ and anthracene-1,2-DCE solution at 24°C forms the radical cation of anthracene with absorption peak at 725 nm. . The radical cation is formed through the charge transfer reaction with solvent radical cation, reaction (33)* Addition of ammonia competes with the formation process and reduces the 79 optical density of anthracene radical cation at 725 nm. The reaction scheme is shown as follows.

DCEt + (§X§X§Ï • (33) DCE^ + KHo ------^ product (3^)

where DCE» represents the solvent radical cation derived from

1,2-DCE.

In the previous discussions, solvent cation was suggested to be the radical cation derived from 1,2-DCE. It reacts with dibenzylmercury to form benzyl cation and is scavenged by ammonia. Let us rewrite reaction (21) and (22).

+ (PhCH^)2Hg 2 ^ PhCH® (21)

^ product (22)

The rate constant has been determined to be 3»^ x 10^

M~^sec~^ or 2.5 x 10^ M“^sec~^ by competition kinetics. If the radical cation of 1,2-DCE, DCE^, is indeed the common precursor species for the formation of benzyl cation from dibenzylmercury and for the formation of anthracene radical cation from anthracene, then should have the same value as kgg.

Figure 32 shows the plot of OI>a„thraoeneV°“*anthraoenet

-1.versus [n h ^J /[anthracen^. A straight line was obtained _ 9 i:m g 6

S g 3 2

iNH^]/[anthracene] x 10

Figure 32. The plot of (OD/OD )-l versus [NHo]/lanthracene]. The optical density of anthracene radical cation was taken at 725 nm. A 80 ns pulse was used. 81 with slope = 0.2. Although the formation rate constant of anthracene radical cation, has not been directly measured, rate constants of several similar electron transfer reactions h 2 have been determined in 1,2-DCE solution such as the forma tion of biphenyl and p-terphenyl radical cations.

DCE^ + biphenyl — biphenyl^ (35)

DCEt + p-terphenyl ----> p-terphenylt (3 6)

The values of k^^ and k^^ were determined to be 1.4 x 10^® and 1 .4 5 X 10^*^ M“^sec~^ respectively. Moreover, the formation rate constants of benzyl cation, k^^, and benzhydryl cation, kg^, have also been determined to be 1.3 x 10^® and 1.4 x 10^®

M”^sec~^ respectively. All of these electron transfer reactions have almost the same rate constants and they are believed to be diffusion controlledi Therefore it is very reasonable to assume the value of k^^ to be 1.4 x 10^® M~^sec“^. By using this value for k^^ and 0.2 for the slope, k^^ was determined to be 2.8 x 10^ M”^sec~^. This is in excellent agreement with the value of k^g, 3*4 x 10^ or 2.5 x 10^

M~^sec ^ . On the basis of all the experimental data obtained so far, it is quite certain that the radical cation of 1,2-DCE is the species that is responsible for the formation of aromatic radical cation from parent aromatic molecule, for the formation of benzyl cation from dibenzylmercury, and for the formation of 82 part of the benzhydryl cation from benzhydryl bromide. The

other solvent cation that responsible for the formation of

benzhydryl and trityl cations from alcohol and bromide com

pounds is believed to be a solvent chloroethyl cation

The following experiments were designed to demonstrate the

existence of such ions in irradiated 1,2-DCE solution.

Titration experiments

e It is known that alkyl carbénium ions, such as ClCHpCHp

derived from 1,2-DCE, do not have absorptions above 220 nm .

Indirect method has to be used to detect its existence. One way to achieve this is to "titrate" the ion by a reaction that will yield a unique and observable product. The formation

of anthracene radical cation is an example of this method.

The detection of anthracene radical cation formed by the

electron transfer reaction, reaction (33), established the

existence of solvent radical cation.

In order to titrate carbénium ion, the best choice seems

to be alkene compounds. The fact that radical cation and

carbénium ion react with alkene compound to yield different products makes the experiment feasible. Two alkene compounds, 1 ,1-diphenylethylene and 2,4-dimethyl-l,3-pentadiene were used. The results will be presented as follows.

(1) Radiolysis of 1 ,1-diphenylethylene-DCE solution. 83

Reaction between solvent radical cation and 1 ,1-diphenyl ethylene forms the radical cation of 1 ,1-diphenylethylene, reaction (37), due to the difference between the ionization potential of 1,1-diphenylethylene (<8.86 ev and 1,2-DCE

(11.05 ev^^). Reaction between solvent chloroethyl cation and 1 ,1-diphenylethylene proceeds via the proton transfer or condensation process and forms a more stable carbénium ion of benzhydryl type, reaction (38).

DCEt + PhgC=CH2 --- > DCE + (37) (38)

where DCE* = radical cation derived from 1,2-DCE.

The product ions of reaction (37) and (38), a radical cation and a carbénium ion of 1 ,1-diphenylethylene, have distinctly different absorption spectrum. The radical cation of 1 ,1-diphenylethylene has been formed in the radiolysis of

1 ,1-diphenylethylene-sec-butyl chloride matrix at 77°K^^. It absorbs at about 1190 nm and 390-400 nm. The methyl benzhydryl, ethyl benzhydryl, and isopropyl benzhydryl cations have been generated in concentrated sulfuric acid and super acid (FSO^H-

SbF^) solution-^®. They all absorb in the 422-430 nm region.

Therefore, the identification of carbénium ion of benzhydryl 84 type in the irradiated 1 ,1-diphenylethylene-DCE solution will

indicate the existence of solvent chloroethyl cation, DCsP.

The time dependent spectrum of irradiated 1 ,1-diphenyl

ethyl ene-DCE solution is shown in Figure 33. Three transient

species appear in the region above 330 nm. The transient that

absorbs at 340 nm is scavenged by oxygen and should be a free

radical of benzhydryl type. The transient that absorbs at 390-400 nm has a half-life time about 2 microsecond and may

be identified with the radical cation of 1 ,1-diphenylethylene

based on the following reasons ; (a) Its absorption maximum

is similar to that of the reported spectrum of 1 ,1-diphenyl

ethylene radical cation^"^. (b) It is not scavenged by oxygen.

This suggests that it is a cationic species, (c) It is not

scavenged by ammonia which is a typical cationic scavenger.

This is best explained by the difference of ionization potentials between 1 ,1-diphenylethylene, <8.86ev^^, and

ammonia, 10.34ev^^.

The transient species that absorbs at about 430 nm has a half-life time about 30-40 microsecond. It is believed to

be a carbénium ion of benzhydryl type based on the following

reasons: (a) The spectrum of this species after correction

for residual absorption is shown in Figure 34(a). The

position and half-width of the absorption peak is similar to

that of the reported spectra of carbénium ions of benzhydryl

type as shown in Figure 34(b). (b) The transient species

is not scavenged by oxygen. This indicates that it is a 85

12 « 10 rt=0 .2 ps ' S:^=2 ws / \ d : = 5 ps

3 2 0 330 340 350 360 370 38O 390 400 410 420 430 440 450 460 Wavelength, nm

Figure 33- The time dependent spectrum resulting from irra diation of a 0.0 3 9 M solution of 1 ,1-diphenylethylene in 1,2-DCE under 1 atm air. A 80 ns pulse was used. 86

(a) 4

0 180 220 260 300 340 380 420 460 500

422

(b) X vu

Wavelength, nm

Figure 34. (a) The optical absorption spectrum of the substituted henzhydryl cation from Figure 33- Solid and open circles are from two different experimental.runs. (b) The optical absorption spectra of two henzhydryl type carbénium ions. The solid curve is due to the diphenylmethyl carbénium ion. The broken curve is due to the diphenylethyl carbénium ion. Both spectra are from reference 50* 87 cationic species, (c) It is scavenged by ammonia. Figure

35 shows the plot of pseudo-first order rate constant versus concentration of ammonia. A straight line was obtained with second order rate constant equal to 9-3 x 10^ M~^sec~^. This is consistent with the rate constants of benzyl, henzhydryl, and trityl cation with ammonia which were obtained previously'.25

These rate constants are shown in Table 2.

Table 2. Rate constants between several carbénium ions and ammonia at 24 C.

Reaction

4.2 X 10^

4.3 X 10^

PhgC-CHgR + NH^ 9.3 X 10® 2.4 X lo"^ Ph^C® + NH^

1.5 X 10^ CH o CH..R

(d) The yield curve of this transient species, a substituted henzhydryl cation, has also been determined as shown in Figure

36. Figure 37 shows the decay curve of this ion at 430 nm.

The initial fast decay is due to the overlapping absorption of the 1,1-diphenylethylene radical cation . It disappears within about 1 microsecond. The optical density taken 1 micro second after the pulse is approximately equal to the optical density at time=0 due to the long lifetime of the henzhydryl ?

Figure 3 5. The plot of pseudo-first order rate constant versus the concentration of ammonia for the reaction between substituted henzhydryl cation and ammonia in 1,2-DCE at 24°G . 6

5 4

Figure 3 6. The plot of optical density at 430 nm as a function of the concentration of diphenylethylene in the irradiated diphenylethylene-DCE solution. A 80 ns pulse was used. The points were taken 1 ps after the pulse. The dashed line indicates the plateau. 90

50 mv]

2 jis

Figure 37. The decay curve of substituted henzhydryl cation at 430 nm in the irradiated diphenylethylene-DCE solution with [diphenylethylene] = O.O39 M. A 80 ns pulse was used. The initial fast decay is due to the overlapping absorption of the diphenylethylene radical cation. 91

cation (t^y^ = 30-40 microsecond). From Figure 36 it is

apparent that a plateau is reached for concentration higher

than about 4 x 10“^ . By using the average of the extinction

coefficients of methyl henzhydryl cation and ethyl henzhydryl

cation, 33000 M~^cm~^, the concentration of the solvent chloro-

ethyl cation responsible for the formation of this substituted

henzhydryl cation was calculated to be 3.4 x . This is

shown in Figure 23 and is in excellent agreement with the

previously determined value. Although the exact structure of the substituted henzhydryl cation formed in the radiolysis

can not be decided due to the similarity of absorption spectra,

between these henzhydryl cations, the extinction coefficient

is expected to be close to the value that was used, 33000 M“^

cm“^, with an uncertainty about ± 20^.

The result of this foregoing titration experiment gives another piece of supportive evidence about the existence of

solvent chioroethyl cation.

(2) Radiolysis of 2,4-dimethyl-l,3-pentadiene (DMPD)-l,2-DCE

solution.

The existence of solvent chloroethyl cation may also be

demonstrated by titrating it with DMPD according to the

following reaction:

„ CH, ‘f”3 CH, H CH, DCE® 4. c H b c ^ C H - C .C H g ------> 92 where R = H or CICH^CH^.

This reaction may he either a proton transfer or a condensation reaction. The product ion in both cases is basically a 2,4- dimethylpentenyl carbénium ion which has been formed in sulfuric acid-^^ with absorption peak at 305nm.

The radical cations of 1,3-butadiene and several methyl substituted 1 ,3-butadienes have been formed in the radiolysis of sec-butyl chloride matrix at 77°K-^^. The peaks of the absorption spectra extend from 425 nm for 1 ,3-butadiene to

469 nm for 2,3-dimethyl-l,3-butadiene. It shifts to the longer wavelength as the number of methyl substituents increases.

Therefore, it is reasonable to expect that the radical cation of DMPD would absorb in the longer wavelength region beyond

450 nm and will not interfere with the observation of 2,4- dimethylpentenyl carbénium ion.

The time dependent spectrum of irradiated DMPD-1,2-DCE solution under vacuum is shown in Figure 3 8. Figure 39 shows the spectrum from the same solution saturated with oxygen.

The transient species that was scavenged by oxygen has a major absorption peak at about 280 nm and a smaller peak at about

435 nm which is the typical absorption spectrum of allylic type free radical. The transient that absorbs at about

307-313 nm is believed to be the substituted 2,4-dimethylpentenyl carbénium ion based on the following reasons: (a)

The absorption peak is close to the reported value of 2,4- dimethylpentenyl carbénium ion in sulfuric acid^^, 305 nm. 93

12 o t=0.4 • t=4 jis A t= end of pulse

X g

270 300 330 360 390 Wavelength, nm

Figure 38. The time dependent spectrum resulting from irradiation of a 0.304 M solution of DMPD in 1,2-DCE (o and • and a 0.245 M solution of DMPD in 1,2-DCE (A) under vacuum. A 80 ns pulse was used in both cases. 94

3 ;=0.4 us 2

0 270 280 290 300 310 320 330 340 350 360 370 Wavelength, nm

Figure 39. The time dependent spectrum resulting from irradiation of a 0.304 M solution of DMPD in 1,2-DCE saturated with oxygen. A 80 ns pulse was used. 95

(Id) It is not scavenged by oxygen. This suggests that it is a cationic species, (c) It is not a radical cation of DMPD because the radical cation of DMPD is expected to absorb in the longer wavelength region beyond 450 nm. (d) It is scavenged by ammonia. The plot of pseudo-first order rate constant versus concentration of ammonia is shown in Figure 40. A straight line was obtained with second order rate constant equal to

1.5 X 10^ M “^sec“^. This is listed in Table 2 and is consis tent with the rest of the data, (e) The reaction between henzhydryl cation and DMPD also forms a transient species absorbing at the same wavelength. This is due to the reaction

(40).

CH, @ CH, 1 ^ CH, p CH, PhgOH® + (,jjbc=CH-C=CH2 — *

The details of this experiment will be presented later. At present, this is sufficient to show that the transient species absorbing at about 307-313 nm is a carbénium ion of 2,4- dimethyl-pentenyl type. It also indicates the existence of

chloroethyl cation.

Proposed mechanism

The above experimental data firmly demonstrates that the radiolysis of 1,2-DCE molecules forms two types of cationic species; radical cations and chloroethyl cations, They have 2

0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 15 16 17 18 19 20 [nH^I x 10^, M Figure 40. The plot of pseudo-first order rate constant versus the concentration of ammonia for the reaction between tetra-substituted allyl cation and ammonia in 1,2-DCE solution at 24°C. 97 different reactivity toward ammonia and can "be selectively scavenged by various precursor compounds to form various arylcarbénium ions.

The radical cations may be the parent radical cations ;

ClCH^CHgCl^, or the fragmented radical cations, CICHCH^'. In U7 the gas phase it has been shown that the loss of HCl from parent radical cation is the lowest energy process. Both

ClCHgCHgCl* and CICHCH^^ were formed by the electron impact of 1,2-DCE in gas phase. In solid phase^^, the radiolysis of n-butyl chloride matrix at 77°K also forms a parent radical cation, Cj^H^Cl^, and a fragmented radical cation, C^Hgt, through the loss of HCl. Similar process of HCl loss from parent radical cation of alkyl chloride molecule is common in the mass spectrometry study^^'^^'-^^. Therefore, it is possible that both parent radical cations and fragmented radical cations, C1CH2CH2C1^ and CICHCH^I^, are produced in the radiolysis of 1,2-DCE in liquid phase. They react with appropriate aromatic solutes to form aromatic radical cations through non-dissociative charge transfer process or form arylcarbénium ions through dissociative charge transfer process.

The experimental results presented in this work show that, if indeed both radical cations are formed, they are kinetically indistinguishable and must have similar rate constants for charge transfer reactions. This is consistent with the observation made my Arai, Kira, and Imamura^^. In their study they found that both parent radical cation and fragmented 98 radical cation of n-butyl chloride, and react with biphenyl to form radical cation of biphenyl at a similar rate at 133°K^^. The other cationic species formed in the radiolysis of

formed through the loss of Cl atom from parent radical cation.

There are several structural possibilities for this ion^"^: it may be a static symmetrically bridged ion (I), a static unsymmetrically bridged ion (II), or a chloronium ion in equilibrium with a pair of open-chain 2-chlorocarbénium ions

(III).

(X) (XX, 01 Cl

(III) CH^CH.Cl^CH,— CHp;=± ClCHpCH^ ® '^Cl

The carbon-13 nmr studies show that the structure of the chloroethyl cation is that of the symmetrically bridged ion(I) in SOgClF solution at -80°C. The structure (III) is energeti cally unfavorable in this case because the open ion is a primary carbénium ion. Several theoretical calculations^^,62,63 also show that the cyclic ethylene chloronium ion is more stable than the open chain carbénium ion by about 9-19 kcal/ mole depending on the calculation method used^^. 99 Ionization and fragmentation of 1,2-DCE molecule has also been studied in the gas phase by photoionization mass ^7- spectrometry and ion cyclotron resonance spectroscopy

Ethyl en e chloronium ion, parent radical cation C1CH2CH2C1^,'

is certainly in agreement with the conclusion made in this study.

In the study of the ion-molecule reactions of ethyl chloride by high-pressure mass spectrometer-^^Luczynski et al found that, in the pressure range of 0 .01-1 torr, most of the parent radical cation of ethyl chloride, C^H^Clt, formed by irradiation or photoionization, react with the parent molecules to give C^H^CIH® species, which in turn condense with a second parent molecule with HCl elimination, thus forming dialkylchloronium ions, RCIR.

CgH^ci: (C2H^)C1(G2H^) (41)

Similar results have also been obtained by other workers^^’^*^.

This type of reactions is possible to occur in the radiolysis of pure 1,2-DCE liquid. The possibility that the protonated

1,2-DCE, (C1CH2CH2C1)H®, instead of ethylene chloronium ion, may be identified to the solvent cation in this study can be ruled out on the following bases; (a) Let us assume that the protonated 1,2-DCE is indeed the solvent cation S^, then the reaction scheme of the formation of benzyl and henzhydryl cation from dibenzylmercury and henzhydryl bromide can be written as;

(42) Ph^CHBr (DCE)H Ph^CH

From this reaction scheme, the concentration of benzyl cation at the plateau of the yield curve should equal to that of henzhydryl cation. This is in contrast to the experimental results, (b) If is assumed to be (DCE)H®, then the reaction @ A ® scheme of the formation of Ph^C , Ph^CH^, and Ph^C-CH^R from

Ph^COH, PhgCHOH, and Ph2C=CH2 respectively can be written as:

Ph^COH DCE. products DCE (43)

(DCE)H®

According to this reaction scheme, as the concentration of solute increases the concentration of carbénium ion may increase initially but will drop eventually due to the competition between solute and DCE for DCE*. This is also in contrast to the experimental results. Therefore, the solvent 101

cation should not be identified as the protonated 1,2-DCE.

An explanation that can explain both the results in this work and the results obtained by the high-pressure mass

spectrometer is as follows; In the pure 1,2-DCE liquid,

radical cation of 1,2-DCE can abstract a H atom from parent

1,2-DCE molecule to form protonated 1,2-DCE, which in turn may condense with another 1,2-DCE to form a dialkylchloronium

ion as observed in the high-pressure mass spectrometry study.

However, in the presence of high concentration of aromatic

solutes, the reaction between radical cations of 1,2-DCE

and solutes becomes the dominant process. The concentration of protonated 1,2-DCE becomes negligible and therefore is not responsible for the formation of arylcarbénium ions.

It is worthwhile to mention the work done by Arai, Kira, and Imamura about the mechanism of the formation of radical

cations in the radiolysis of butyl chloride matrix at low

temperatures^^' . In their study the parent radical cation

of n-butyl chloride, C^H^Cl"^, and the fragmented radical

cation, Cj^Hg*, with absorption maxima at 550 nm and 450 nm

respectively, were formed by the radiolysis of n-butyl chloride.

Irradiation of biphenyl-n-butyl chloride solution formed biphenyl radical cation with absorption maximum at 700 nm.

However, they found that at 100°K all the biphenyl radical

cations were formed simultaneously with pulse duration. The

decay rate of radical cations C2^H^Clt and C^^Hg» were unchanged by the addition of biphenyl. This indicates that the biphenyl 102 radical cations were not formed "by charge transfer reaction between solvent radical cation and "biphenyl at 100°K. Addition of biphenyl to n-butyl chloride at 77°K reduces the formation of both radical cations, and by the same ratio, and is accompanied by the formation of biphenyl radical cations. When the temperature was raised to 133°K, a portion of the biphenyl radical cation was found to be formed by the charge transfer process between solvent radical cation and biphenyl, and there is still a significant amount produced simultaneously with pulse irradiation.

From the above observations, they concluded that radical cations Cj^H^Cl^ and must originate from a common pre cursor, which reacts with biphenyl forming its radical cation at low temperature. The nature of the precursor is believed to be a vibrationally excited n-butyl chloride radical cation. The reaction scheme is as follows:

► (n-Cj^H^Cl* )* -T--- > n-Cji^H^Clt

'----> Cji^Hg. + HCl (44)

— » W -

Since the molecular diffusion is prohibited in the solid phase, the migration of positive charge from excited solvent radical cation to solute must go through the so called

"resonance charge transfer" or "electron jump" process. The excited radical cations transfer positive charges and energy 103

to neighboring neutral molecules. The migration is a succession

of such transfers. During migration, vibrationally excited n-butyl chloride radical cations either decompose into radical

cation , or transfer charges to solute molecules on

encounters. When the excited ions are deactivated below a

certain level, the ions are then trapped in a solid or liquid, and/or are then solvated. As temperature is increased to 133°K

these "stabilized" solvent radical cations then react with biphenyl to form its radical cation by diffusive process, although some portion of the biphenyl cation are still formed by the non-diffusive process.

On the basis of their result, one would like to know what kind of role this excited radical cation plays in the radiolysis of 1,2-DCE liquid at 24°C.

It has been suggested^^ that non-diffusive processes were responsible for the formation of aromatic radical cations in 1,2-DCE at 24°C based on the fact that their formation were too fast to be observed at that time, and an estimation gave the value 3 x 10^^M~^sec~^ as the lower limit of the rate constants which was too fast to be accounted for by molecular diffusion^^. However, subsequent studies^^ with better resolu tion demonstrated that the formation rate curves of these aromatic radical cations can be observed, and their formation rates, typically 1 .4 x 10^^M~^sec~^, can be accounted for by the diffusion process. The formation rate constants of arylcarbénium ions have also been.determined^^’^^, and all of them can be explained by the diffusion process. Figure 4l 104 and Figure 42 show the formation rate curves of p-terphenyl radical cation and henzhydryl carbénium ion in 1,2-DCE at 24°0.

In Figure 4l (taken from reference 42), the formation rate curve of p-terphenyl radical cation followed first order kinetics. The first order rate constants obtained were linear in solute concentration, indicating the normal bimolecular process. No abrupt change was observed on the curve. The rate curve can extrapolated back to the beginning of the pulse, thus demonstrating no other fast process was occuring during the pulse. In Figure 42(a), an abrupt increase in optical density was observed after the pulse on the formation rate curve of henzhydryl cation. This is due to the solvent absorption as shown in Figure 42(b) which was taken in the pure 1,2-DCE at the same wavelength. Therefore, no fast process was observed either, the rate curve of henzhydryl cation can also be explained by the normal bimolecular diffusion process. In summary, there is no need to invoke the idea of non-diffusive process in the radiolysis of 1 ,2-

DCE at 24°G.

The above conclusion is parallel to the findings of

Arai et al^^ in some way. They found that in the n-butyl chloride solid at 100°K all the biphenyl radical cations were formed by the non-diffusive process. As the temperature was raised to 133°K when the diffusion process is allowable, a less; amount of radical cations was formed by the non-diffusive process and some of the radical cations were formed by the 105

20 mv

200 ns

Figure . The formation rate curve of p-terphenyl radical cation monitored at 96O nm. This Figure is taken from reference 42. 106

50 mv

500 ns

50 mv

500 ns

Figure 42. (a) The formation rate curve of henzhydryl cation in the irradiated henzhydryl hromide-DCE solution. The wavelength monitored was 449 nm. (h) The optical absorption at 449 nm in the irradiated pure 1,2-DCE solution. 80 ns pulses were used in hoth cases. 107 slower diffusive process. It is not surprising that when the temperature is raised to all the radical cations are formed by the diffusive process.

In summary, it seems that in the 1,2-DCE liquid at 24°C, the excited radical cation of 1,2-DCE can he deactivated much faster than it can be in the solid at 1G0°K, thus preventing the charge transer process from occuring between excited radical cations and solutes. Bimolecular diffusion process between

"stablized" radical cations and solutes is the dominant reaction path for the formation of cationic species in 1,2-DCE at 24°C.

Based on all the information obtained and discussed above, the formation of cationic species, aromatic radical cations or arylcarbenium ions, in 1,2-DCE can be represented by the following general mechanism.

ClCHgCHgCl ClCHgCHgClt )* ■

' CH^CHp + Cl- \ l

in the solid phase,

(ClCHgCH^Clt)* + ClCHgCHgCl -- » ClCH^CH^Cl + (ClCHgCHgClT)* (46) Am — *> cicHgCHgCl + Amt (4?)

in the pure 1,2-DCE liquid. 108

(48)

+ HCl (49)

in the 1,2-DCE solution of aromatic solutes, DCE» + Am ----- > DCE + Amt (5 0)

DCEt + (PhCH2)gHg » PhCHg® + PhCH^Hg.-t- DCE (51)

DCE' ---- 1 + Ph^CBr > Ph_C^ (52) CH 2 CH 2 — I 3 3 Cl

DCE' ---- 1 0 + Ph.CHBr > Ph^Cir (53) c h ^c h ^--- 1 Cl

CHpCHp @ ^ + Ph^COH > Ph^C® (54)

CHpCHp a ^C1 "■ » Ph^CH® (55)

Ph,C-CH-5 + CICHCH^

(56)

where DCE' = ClCH^CHgCl' and CICHCH^. Am = Aromatic molecules such as biphenyl, p-terphenyl, and anthracene. 109

II. REACTIONS BETWEEN ARYLCARBENIUM IONS AND ALKENES

The reactivity of various alkenes toward carhenium ions has never been directly determined to the best of author's knowledge. The accumulation of such data is important not only for the understanding of this type of fundamental reaction but also for the study of cationic vinyl polymeri zation reactions. There is a substantial amount of work about the kinetic study of cationic vinyl polymerization in the literature^^”^^^. Many books and review papers have been written about this subject^^”"^-^. Different initiators such as pro tonic acids"^^"^^, iodine^*^, Friedel-Crafts reagents^^”^^, carbocation salts^^”^^, and radiation®"^”^^ have been used to initiate the cationic vinyl polymerization reactions. Fast- reaction techniques such as stopped-flow^^ and pulse radiolysis^^"^^^ have also been used for the study of short lived transient species in the cationic polymerization. The chemically-induced initiation usually introduces complexity into the system. The initiation equilibrium and the equili brium between ion-pairs and free ions have to be considered in the kinetic scheme. The reactions with resulting initiator fragments and aggregation phenomenon of cations also make the experimental data hard to interpret. These complications sometimes may be eliminated by using ionization radiation as initiator. Rate constants purely contributed by free ions may be obtained. For the sake of comparison, the propagation rate constants of the polymerization of styrene using different initiators are listed in Table 3^^. The values of are quite different between chemically-induced and radiation- induced polymerization reactions. A cationic species is believed to be the propagating center in all the cases. The result from Y -radiation is recognized as the reactivity of free ions. Interpretation of the remaining data is still speculative. Such values cannot be representative of a propagating free ion simply because they are numerically too low. It is generally believed that those k^ value obtained by chemically-induced initiation may represent the reactivity of ion-pairs or the statistical averages of free ions and ion-pair contributions. In short, the above example shows that at the present time there is still a lot of confusion among the existing data.

The generation of arylcarbénium ions by pulse radiolysis method in 1,2-DCE solution allows the direct determination of such rate constants by observing the arylcarbénium ions spectrophotometrically. The method is straightforward and without the complications usually associated with chemically- induced and -radiation-induced polymerization. In the absence of nucleophilic reactant, the arylcarbénium ion recombines with the counterion in the system, chloride ion, which is formed from 1,2-DCE in a dissociative electron attachment process. As the alkene is added, a competition reaction occurs. The reaction scheme is as follows; Table 3. Propagation rate constants for the cationic polymerization of styrene, taken from reference 64.

Initiator Solvent Temp.(°G) kp(mole"^sec“^l)

HCIO^ ClCHgCHgCl 25 1 7 .0

HCIO;^ CCl^ 20 1.2 X 10"^

GlCHgGHgCl 25 7.6 SnCl^ GlGHgGHgGl 30 0.42

GlGHgGHgGl X 10"^ "2 30 3 .5 y-ray bulk 15 3 .5 X 10^ 112

R® + Cl“ ----- ^ product (57)

R® + alkene > product (58)

where R® represents arylcarhenium ion. The concentration of alkene is in excess such that reaction (58) is pseudo-first order and competes overwhemingly over the natural decay, reaction (57)• The slope of the plot of In OD of arylcarhenium ion versus time gives the pseudo-first order rate constant of reaction (5 8). The plot of pseudo-first order rate constant versus concentration of alkene yields the second order rate constant of reaction (5 8) as the slope.

Possible reaction path.

There are three possible reaction paths between an aryl- carbénium ion and an alkene molecule: electron transfer reac tion, condensation reaction, and hydride transfer reaction

(when the alkene molecule has a methyl substituent on the double bond). They are written as follows:

R ® . . u — (%)

R® + -C=C- ---- > -C-C- condensation (6 0)

r '*’ + > RH + transfer ( 6l )

where R represents either a benzyl cation or a benzhydryl 113 cation. The proton transfer from arylcarbénium ion to alkene molecule is not likely to occur because the resulting transient is a highly unstable carbene.

Reaction (59), a electron transfer reaction, yields a free radical and a radical cation of alkene. This reaction path was ruled out in this work based on the following reasons; (a) The ionization potentials of benzyl radical and benzhydryl radical are 7.76 and 7.32 ev respectively. The ionization potentials of some of the alkenes used in this work are^*^^: ethylene-10.5 ev, propylene-9.73 ev, isobutylene -9 .3 5 ev, 1,3-butadiene-9.24 ev, and 2,3-dimethyl-l,3-butadiene

-8 .7 2 ev. All of those alkenes have a higher ionization potential than that of benzyl radical or benzhydryl radical.

Therefore the electron transfer reaction is not likely to occur energetically, (b) Many of the radical cations of alkenes have known absorption spectra obtained by the low temperature organic glass study^^. In this work no such radical cation has been detected. This indicates that the electron transfer reaction channel does not make any signi ficant contribution to the rate constant measured.

Reaction (6 1), a hydride transfer reaction, yields a stable molecule and an allylic carbénium ion. This reaction path was ruled out based on the following two reasons: (a) The hydride transfer reaction between benzyl cation and triphenylmethane was chosen as the reference reaction:

^ PhCH- + Ph_C® (62) 114

The occurence of this reaction can be monitored by both the

decay of benzyl cation at 364 nm and the formation of trityl

cation at 434 nm. When the concentration of triphenylmethane

was increased up to 0.145 M, there was still no significant

change on the decay rate of benzyl cation. No trityl cation was observed either. An upper limit of 5x10^ M“^sec“^ was placed on this rate constant. If the value of the rate

constant were higher than 5x10^ M“^sec“^, then 0.145 M of

triphenylmethane should have made an observable change on the decay curve of benzyl cation. Since trityl cation is typical ly much more stable than the allylic cation generated in reaction (6l), the rate constant of reaction (6l) is expected to be much smaller than that of reaction (62)(i.e. 10^).

As will be shown later, the measured rate constants between arylcarbénium ions and alkenes range from 8.7x10^ to 1.0x10^ M“^sec~^. This clearly indicates that the hydride transfer reaction channel does not make any significant contribution to these rate data, (b) The hydride transfer reaction between arylcarbénium ion and the methyl substituted 1,3-butadiene will yield a dienyl cation,

0 CHp , , — ^RH+ ^^C=C-C=CC (63) which has an absorption maximum typically around 400 nm^^^.

In this work no such dienyl cation was detected for the reaction between benzhydryl cation and various substituted 115

1.3-butadienes. It was concluded that the hydride transfer reaction channel was not important in the present study.

In summary, the rate constants measured in this work by monitoring the decay of arylcarhenium ion should represent the condensation reaction, (60). The product ion of this reaction is either an alkyl carhenium ion ( when the alkene molecule contains only one double hond) or an allylic carhenium ion (when the alkene molecule contains two conjugated double bonds). The alkyl carbénium ions are known to absorb below

220 nm^^ which is in a region not accessible by the present technique. The absorption spectra of allylic carbénium ions are generally unknown except those are tetrasubstituted at the end of the allylic system^^^. The reaction between benzhydryl cation and 2,4-dimethyl-l,3-pentadiene yields a tetrasubstituted allyl cation and was detected in this work. The details will be presented later.

The reactions between arylcarbénium ions and the following alkene compounds were studied in this work; ethylene, propylene, isobutylene, 1,3-butadiene (BD), cyclohexene, 1 ,3-cyclohexadiene,

3.3-dimethyl-l-butene (neohexene), allylbenzene, 2-methyl-l,3- butadiene (isoprene, MED), 2,3-dimethyl-l,3-butadiene (DMBD),

4-methyl-1 ,3-pentadiene (MFD), 2,4-dimethyl-l,3-pentadiene

(DMPD). They will be divided into several groups and discussed separately. 116

Ethylene, propylene, isobutylene, and cyclohexene

The rate constants between benzyl cation and ethylene,

propylene, isobutylene, and cyclohexene were determined by

monitoring the decay of benzyl cation. As mentioned before,

the decay curve of benzyl cation at 364 nm was overlapped by

an unknown species with absorption maximum at about 323 nm.

This unknown species has a much longer life time than that

of the benzyl cation. Figure 13 and Figure 43 show the time

dependent spectra of irradiated dibenzylmercury-DCE solution

in the presence of propylene and isobutylene respectively.

The typical decay curves of benzyl cation in the presence of

propylene, isobutylene, and cyclohexene are shown in Figure

l4 and Figure 44. The slow decay due to the unknown species

can be treated as a plateau relative to the fast decay of benzyl cation. The kinetic equations are solved in

Appendix C. By plotting -In (OD-OI^) versus time, pseudo-

first order rate constant was obtained from the slope. The

plots of pseudo-first order rate constants versus concentra

tion of alkenes are shown in Figure 45, Figure 46, and Figure

47. The slopes of the plots give second order rate constants

which are listed in Table 4. The rate constant for ethylene was estimated to be smaller than 10^ M'^sec"^ based on the fact that the decay rate of benzyl cation remained unchanged when the concentration of ethylene was increased up to 0.211 M. From Table 4 it is clear that the trend of rate t=end of pulse

Wavelength, nm

Figure 43. The time dependent spectrum resulting from irradiation of a isohutylene- dibenzylmercury-1,2-DGE solution under vacuum. The concentration of isobutylene is 0.2 M. The concentration of dibenzylmercury is 5*6x10 exp(-3) M. A 80 ns pulse was used. ^ 118

(a)

50 mv

500 ns

706 mv (b) 50 mv

1 )XS

Figure 44 (a). The decay curve of benzyl cation monitored at 359 nm in the presence of O.O97 M isobutylene. A 80 ns pulse was used.

Figure 44 (b). The decay curve of benzyl cation monitored at 364 nm in the presence of 0.055 M cyclohexene. A 80 ns pulse was used. 2

0 G1 2 3 5 6 7 8 9 10 [Propenejx 10, M

Figure 4$. The plot of pseudo-first order rate constants versus the concentrations of propene for the reaction between benzyl cation and propene in 1,2-DCE at 24 0. 120

X 1, 20

10 20 [isobutylene] x 10^, M

Figure 46. The plot of pseudo-first order rate constant versus concentration of isohutylene for the reaction "between benzyl cation and isobutylene in 1,2-DCE at 24 C. 121

[1,3-Butadiene] x 10, M

Figure 47. The plot of pseudo-first order rate constant versus the concentration of 1 ,3-hutadiene for the reaction between benzyl cation and 1,3-butadiene in 1,2-DCE at 24 0. Table 4. Rate constants for reactions between benzyl cation and alkenes in 1,2-DCE at 24 G.

ethylene propylene isobutylene cyclohexene CHg=CHg CH^-CH=CHg C H p 0

PhCH® <10^ 1.9 X 10^ 1.9 X lo"^ 9.4 X 10^ ill 5%) ill5fo) illSfo)

CH„ product CHpCHpR CHo-CH-CH.R ion Q CHj-C-OHgR C r ” ® (1°) (2°) (3°) (2°)

R represents PhCH2" 123

constants increases from ethylene, propylene, cyclohexene,

to isobutylene. This finding is not surprising. It is known among polymer chemists that isobutylene can be easily poly merized through cationic mechanism whereas ethylene cannot be polymerized under normal condition (i.e. without strong

catalyst, high pressure, etc.). This trend can be explained on the basis of the stability of the product ion. As can be seen from Table 4, the increase of rate constant is parallel to the increase of stability of product ion. The product ion from benzyl cation and ethylene is a primary carbénium ion, whereas the product ion from benzyl cation and isobutylene is a much more stable tertiary ion. In other words, the more stable the product ion is, the faster the reaction rate is. This immediately suggests the validity of a crude, but highly successful principle in physical organic chemistry, the linear free energy relation ship (or the product stability principle)The linear free energy relationship says that for the following reaction;

R® + -Ç=Ç- > -Ç-Ç- (60)

the change in activation energy brought about by a change in alkene structure should be proportional to the change in free energy difference of reaction (60),AG (note that

AG is negative for an exothermic reaction such as (60)). If 124

the preexponential factor in the Arrhenius equation is

assumed to be the same for similar reactions (or its

logarithm is a linear function of AG ) , then log k should be

proportional to AG. Since entropy change is usually small

for such reactions, the linear free energy (LFE) principle

thus predicts a linear relationship between log k and heat

of reaction,AH.

The A H of reaction (60) is not available in the liquid

phase. It was estimated from gas phase ionization potentials

and bond energies data. This is justified since both reactant

and product of reaction (60) are ions, their solvation

energies should be approximately the same under normal

circumstances. Appendix D shows the details of the

calculations of AH's for reactions of benzyl cation with

ethylene, propylene, isobutylene, and cyclohexene. When

gas phase ionization potentials and bond energies were not

available, they were estimated as explained in Appendix D.

The plot of log k versus AH is shown in Figure 48. A

straight line was obtained which indicates the validity of

linear free energy relationship in the range we examined.

Since the calculation of AH involves several approxima

tions (see Appendix D), a different approach to estimate the

AH is desired to test the LFE relationship. Burstall has

calculated^the gas phase carbénium ion affinities of

various alkenes including ethylene, propylene, and isobutylene.

The carbénium ions he used include methyl, ethyl, n-propyl, 8

isobutylene cyclohexene 7

propylene M bD 6 1 ,3-butadiene

4 20 12 8 4 04■8 -12 -1616 -20 AH, Kcal/mole

Figure 48. The plot of log k versusAiH for reactions between benzyl cation and alkenes. Calculations of AH's are shown in Appendix D. 126

isopropyl, n-butyl, sec-butyl, and t-butyl cation. The

result of his calculation is shown in Table 5* It can be

seen that for two different carbénium ions AH 's for all

alkenes differ approximately by a constant value (e.g. the

difference between methyl cation and ethyl cation is about

32 kcal/mole for all alkenes). In this work benzyl cation

was used; its difference in AH's with any carbénium ion in

Table 5 should also be a constant value for all alkenes.

Therefore, by using Burstall's AH 's in the LFE plot of benzyl

cation, a straight line with the same slope as in Figure 48

but different intercept should be obtained, if the previous

calculation in Appendix D is correct. Such a plot is shown

in Figure 49. For each carbénium ion AH ’s for all alkenes

were shifted by the difference between its AH for isobutylene

and that of benzyl cation. A reasonable fit was obtained

in spite of so many diverse data. This indicates the

validity of the treatment used in Appendix D and LFE

relationship.

It is also interesting to note that by using Burstall's

AH data for ethylene, the LFE relationship predicts the rate

constant between benzyl cation and ethylene should be within

the range 10^ to 6 x 10^ M“^sec~^. This is certainly in

agreement with the upper limit estimated from the pulse

■ radiolysis experiment, 10 ^ M“^sec~^. Table 5* Gas phase carbénium ion affinities of olefins, taken from reference 1 0 5. olefin ■ ' Rg,A (kcal/mole)

Olefin Me"*" Et"^ n-Pr"^ i-Pr"^ n-Bu^ s-Bu^ t-Bu

CHg=CH2 -6 9 .5 -3 5 .0 -2 2 .5 -8 .5 -2 5 .0 0 .0 +7 .5

MeCH=CH2 -9 0 .5 -6 1 .0 -4 5 .0 -3 0 .5 -4 7 .0 -2 2 .0 -14.5 Me2C=CH2 -1 0 3 .0 -7 1 .0 -5 6 .0 -42.0 -5 8 .5 -33.0 -24.0 8

isobutylene cyclohexene 7

Kx 6 propylene

4 20 16 12 8G.4 8 -12 -16 -20 AH, Kcal/mole

Figure 49. The plot of log k versus AH for reactions between carbénium ions and alkenes. Solid oiroles(#) are calculated in this work as explained in Figure 48. Crosses(X) are taken from reference 105 and normalized to the value of isobutylene as explained in the text. The rate constant between benzyl cation and ethylene is predicted to be within the range R. ■ 129 Ethylene, propylene, neohexene. and allylbenzene

The reactions between benzyl cation and ethylene, propylene, neohexene, and allylbenzene were studied in order

to understand the effect of substituents (i.e. H, CH^, (CH^)^C, and PhCHg) on the rate of condensation reaction. It also indicates the stabilization effect of these substituents on

the neighboring carbon cation if LFE relationship is assumed

to be valid. The product ions of these reactions are:

H CH, (CH_)_C PhCHp 'CH-CH^R, -^^CH-CHgR, ^ ^ CH-CH^R, ^''CH-CH^R e e 0 ® where R represents benzyl cation.

The rate constants for ethylene and propylene have been presented before. Similar procedure was used to determine the rate constants for neohexene and allylbenzene. The time dependent spectra of irradiated neohexene-dibenzylmercury-DCE solution and allylbenzene-dibenzylmercury-DCE solution are shown in Figure 50 and Figure 5 1. The plots of pseudo-first order rate constants versus concentrations of neohexene and allylbenzene are shown in Figure 52 and Figure 53 respectively.

The second order rate constants are listed in Table 6 for comparison. It can be seen that while the differences between these rate constants are small, a trend still can be established. The rate constants decrease in the following order: 130

t=end of pulse

X §

320 330 340 350 360 370 380 390 400 410 420 430 Wavelength, nm

Figure 50. The time dependent spectrum resulting from irradiation of a neohexene-dihenzylmercinry-l,2-DCE solution under vacuum. The concentration of neohexene is 0.451 M. A 80 ns pulse was used. 131

t=end of pulse

§ t=l.2 ps

X ---

Wavelength, nm

Figure 51* The time dependent spectrum resulting from irradiation of a allyl henzene-dibenzylmercury-l,2-DCE solution under vacuum. The concentration of allyl benzene is 0 .4 3 8 M. A 80 ns pulse was used. 132

0 1 2 3 4 5 [Neohexene] xlO, M

Figure 52. The plot of pseudo-first order rate constant versus concentration of neohexene for the reaction between benzyl cation and neohexene in 1,2-DCE at 24 C. 133

5 4

G G 1 2 3 4 5 [Allyl "benzene]xlG,

Figure 53* The plot of pseudo-first order rate constant versus concentration of allyl "benzene for the reaction "between benzyl cation and allyl benzene in 1,2-DCE at 24°C. Table 6. Rate constants for reactions between benzyl cation and alkenes in 1,2-DCE at 2^ C.

GH_ (CHo )o Q PhCHp CH=CHg ^^CH=CHg ^ CH=CHg '^''CH=CHg

PhCH® < 1 0 ^ 1.9 X 10^ 2.2 X 10^ 1.5 X 10^ (±15%) (±15^) i±15%) 135

(CH^)^C > CH^ > PhCHg > H

By assiAming the validity of LFE relationship, the stabiliza tion effect on the neighboring carbon cation decreases in the same order:

(CH^)^C > CH^ > PhCH2>H

There has been quite a lot of work on the effect of substituents on the stabilities of cations^^^”^^^. The early observation^®^’ that the rates of many reactions of molecules containing alkyl substituents decrease in the so-called Baker-Nathan order:

CH^ > CH^CHg > (CH^)2CH > (CH^)^C

which was in contrast to the "normal inductive order":

CH^ < CH^CH2< (CH^)2CH < (CH^)^C

led to the conclusion that C-H hyperconjugation is more effective than C-C hyperconjugation. The latter work found O that the stabilities of the substituted acetyl cations RC=0

(R = CHj, CH^CH2 , (CH^)2CH and (CH^)^C) follow a Baker-Nathan order in super acid solution, whereas a normal inductive order is observed in gas phase^®®. ^ initio molecular orbital 136 theory has also heen used to study the cations RCH® and

RC®=0 (R= CHj, CH^CHg, (CH^)^CH, and In agreement with gas phase data, the calculations produce the normal inductive order of stability for these cations. This led to the conclusion that (a) the Baker-Nathan order observed in solution must be attributed to solvent effects and (b) the

C-C hyperconjugation is more effective than C-H hyperconjuga tion.

The result obtained in this work by pulse radiolysis method in 1,2-DCE solution follows the normal inductive order:

(CH^)^C > CH^

which is consistent with gas phase data and theoretical calculations but in contrast to the Baker-Nathan order observed in acid solution.

Propylene, 1,3-butadiene, cyclohexene, and 1,3-cyclohexadiene

The rate constants for reactions between benzyl cation and 1,3-butadiene and 1,3-cyclohexadiene were determined by similar procedure described earlier. The time dependent spectra of irradiated 1,3-butadiene-dibenzylmercury-DCE solution and 1 ,3-cyclohexadiene-dibenzylmercury-DCE solution are shown in Figure 5^ and Figure 55. The decay curve of 137

^=end of pulse

fc=2 ps X § X).,

300 310 320 330 340 350 360 370 380 390 400 4l0 420 . Wavelength, nm

Figure 54. The time dependent spectrum resulting from irradiation of a 1 ,3-‘butadiene-dihenzylmercury-l ,2-DCE solution under vacuum. The concentration of 1,3-butadiene is 0.4 5 1 M. The concentration of dibenzylmercury is 5 .9 X 10 exp(-3) M. A 80 ns pulse was used. 10

t=end of pulse

cvj

§ N., t=l.l ps

” - -X t=2.6 ps

260 270 280 290 300 310 320 330 340 350 360 370 380 39O 400 4lO 420 Wavelength, nm Figure 55. The time dependent spectrum resulting from irradiation of a 1 ,3-oyclohexadiene- dibenzylmercury-1,2~DCE solution under vacuum. The concentration of 1,3-cîyclohexadiene ^ is 0.109 M. A 80 ns pulse was used. 139

"benzyl cation in the presence of 1,3-butadiene is shown in

Figure 5é. The plots of pseudo-first order rate constants versus concentrations of 1,3-butadiene and 1,3-cyclohexadiene are shown in Figure 57 and Figure 58 respectively. The second order rate constants are listed in Table 7 for comparison. The product ion for each reaction is also listed in the same table.

It is interesting to see that the reaction between benzyl cation and 1,3-butadiene (yields an allyl cation) has a lower rate constant than the one between benzyl cation and propylene (yields a secondary cation), although the allyl cation is more stable than the secondary cation according to the calculation shown in Appendix D. The data point of 1 ,3-butadiene deviates significantly from the linear free energy plot among mono-olefins as shown in

Figure 48. However, the reactivity of cyclohexene ( a mono olefin) toward benzyl cation is lower than that of 1,3-cyclo hexadiene ( a conjugated diene) as expected by the product stability principle. These observations may be explained in the following.

A generalization called the principle of least motion^^"^ is often used by physical organic chemists, together with the product stability principle and asynchronization effect^^"^, to understand the structural effects on reactivity. The principle of least motion, in the words of Rice and

Teller^^^'^^^, states that "those elementary reactions will 140

792 mv

20 mv

Figure 5 6. The decay curve of benzyl cation monitored at 364 nm in the presence of 0.72? M 1,3-butadiene. A 80 ns pulse was used. I4l

[Cyclohexene] xlO^, M

Figure 57. The plot of pseudo-first order rate constant versus concentration of cyclohexene for the reaction between benzyl cation and cyclohexene in 1,2-DCE at 2 k C. 142

0 0 1 2 3 4 56 78 9 10 [1f3-Cyclohexadiene]x 10^

Figure 58. The plot of pseudo-first order rate constant versus concentration of 1,3-cyclohexadiene for the reaction between benzyl cation and 1,3-cyclohexadiene in 1,2-DCE at 24° C. Table 7. Rate constants for reactions between benzyl cation and alkenes in 1,2-DCE at 24°0.

CH^-CH=CHg CHg=CH-CH=GH2 0 0

PhCH® 1.9 X 10^ 8.7 X 10^ 9 . 4 X 10^ 2.7 X lo"^ (+15f.) (± 2 5 ^ ) (+l5fo) (±15^)

CH^=CH-CH-CH^R (2°) ^ product ® 1 O ' " « ° » ion CHo-CH-CHpR C r " (2°) CH^-CH=CH-CH^R (1°) (2°) . C r ' » • >

Units of these rate constants are M“ sec” . R represents PhCHg. 1 # be favored that involve the least change in atomic position and electronic configuration". This principle may be divided into two parts, the principle of least nuclear motion (PLNM), and the principle of least change in electronic configuration. The effect of PLNM may be brought about by the changes in bond lengths and bond angles. It is b e l i e v e d ^ t h a t the sum of the squares of the changes in bond numbers, which is approximately proportional to the sum of the squares of the changes in bond lengths, is a crude measure of that part of the PLNM effect arising from changes in bond lengths.

In the present case, the formation of an allylic cation is accompanied by the shortening of the carbon-carbon single bond and the lengthening of the carbon-carbon double bond as each becomes a bond and a half (i.e. the bond number is

1.5). e CH2=CH-CH=CH2 CHg^CHWDH-.CHgCHgPh bond numbers 2 1 2 1.5 1.5 1

Such a reaction involves (2-1.5)^ + (1-1.5)^ + (2-1)^ or

1.5 in the squares of the changes in bond numbers. On the other hand, the formation of secondary cation involes only a total of 1 in the squares of the changes in bond numbers.

CHo CHq ^"CH=GH2 ■^•^CH-CH2CH2Ph

bond numbers 1^5 This difference of 0.5 in the squares of the changes in bond numbers makes the reaction between benzyl cation and propylene more favorable in terms of PLNM effect. The total estimated PLNM effect could be considerably larger than this if allowance were made for changes in bond angles and for loss of resonance stabilization of the transition state^^^.

Simi

H atom abstraction reactions between methyl radical and propane (yields an isopropyl radical) and propylene (yields an allyl radical). In spite of the greater stability of allyl radical, its formation rate is slower than that of isopropyl radical. This was also explained by the effect of PLNM.

In the cyclic system, the reaction between benzyl cation and cyclohexene yields a cyclic secondary cation, while the reaction between benzyl cation and 1,3-cyclohexadiene yields a cyclic allyl cation. Although the effect of PLNM favors the former reaction, it is obvious that the effect of product stability principle, which favors the latter reaction, is dominant in this case. This is because that the cyclic allyl cation is a very stable species ; its resonance hybrid consists of two cyclic, secondary, and allylic structures whereas the resonance hybrid of the allyl cation produced by the reaction between benzyl cation and 1,3-butadiene consists of one secondary allylic and one primary allylic structures. ^CHgPh CH2=CH-CH-CH2CH2Ph (2°,allyl) ^ (2°,allyl)

CH2-CH=CH-CH2CH2Ph (1°, allyl) ^ (2°, allyl)

The quantitative separation of the effects of product stability principle and PLM, unfortunately, is not feasible at present.

1 .3-Butadiene(BD), 2-methyl-l,3-butadiene(MBD), 2,3-dimeThÿl-

1.3-butadiene (DMBD) , ^-methyl-1,3-pentadiene(BdPD), 2,4-dimethyl-

1.3-pentadi ene(DMPD)

In this section the effects of methyl substituent on the

reactivity of several 1,3-conjugated dienes toward benzhydryl

cation are systematically studied. The product ion for the

reaction between benzhydryl cation and DMPD has also been

identified. They will be discussed separately in the following.

(1) Product ions.

The reaction between benzhydryl cation and DMPD yields

a tetrasubstituted allyl carbénium ion;

@ CH. CH. .CH PhgCH + ^3)c=CH_C=CH2 --- > CH^::C<-^::

which is the only acyclic allyl cation that has a known UV spectrum^ C3,53 ^

Figure 59 shows the time dependent spectra of irradiated 10

t=end of pulse t=60A ns X §

t=540 ns, \ under .A-- —

300 400 500 Wavelength, nm

Figure 59. The time dependent spectra resulting from irradiation of DMPD-benzhydryl bromide- 1, 2-DCE solutions under vacuum (•, o ,a ) and under oxygen (□). The concentrations of DMPD are 0.00831 M and O.OI5I M respectively. 80 ns pulses were used. ^ 148

DMPD-benzhydryl ‘bromide-DCE solution under vacuum and under oxygen. Three transient species can he seen in the spectra.

The species with an absorption peak at 449 nm is benzhydryl cation which decays through reaction (40). The species with an absorption peak at about 328 nm is believed to be benzhy dryl radical which was scavenged by oxygen as shown in the figure. The species with an absorption peak at about 312 nm is believed to be the product ion of reaction (40) based on the following reasons : (a) The absorption peak is close to the reported value of 2,4-dimethylpentenyl carbénium ion in sulfuric acid-^^’^*^^, 305 nm. (b) This species is not scaven ged by oxygen, (c) The reaction between solvent carbénium ion with DMPD, reaction (39), also forms a transient species absorbing at about the same wavelength. This transient species has been shown to be a tetrasubstituted ally cation in the previous section, (d) The formation rate of this species is approximately equal to the decay rate of benzhydryl cation. The details of the rate measurement will be presented later. The dependence of the formation rate on the concentra tion of DMPD excludes the possibility that this transient is a cyclic cation resulting from the unimolecular cyclization of linear allyl cation.

In summary, the transient species with absorption peak at about 312 nm can be identified as the product ion of reac tion (40). This is the first time that both the reactant ion and product ion are observed and identified in the carbénium 149 ion work studied by pulse radiolysis. It clearly demonstrated that the condensation reaction is the main reaction channel for arylcarbénium ions studied in this work. A similar attempt has been made to observe the product ion between benzhydryl cation and MPD:

<0 CHo CHo Pfj u PV» " -"

The product ion of reaction (64) is a trisubstituted allyl carbénium ion which has never been observed before. The time dependent spectra of irradiated MPD-benzhydryl bromide-DCE solution under vacuum and under oxygen are shown in Figure 60.

Similar results as in the case of DMPD were obtained. A transient species with an absorption peak at about 312 nm was separated from the benzhydryl radical by scavenging the latter with oxygen. If this species is indeed the trisubstituted allyl cation, it should also be formed by the reaction between a solvent cation and MPD:

DO^ + ®2>C=CH-CH=CH2 — > ^«CoH^DCE

The spectra of irradiated MPD-DCE solution under vacuum and under ammonia are shown in Figure 6l. It can be seen that a transient species absorbing at around 320-330 nm was scavenged by ammonia. The difference spectrum, obtained by normalizing the two spectra at 271 nm and taking their difference, gives 7 t=end of pulse, under vacuum t=0,6 ps, under vacuum 6 t=5.2 ps, under vacuum t=0.6 ps, under oxygen 5

X § 3

1 0 300 350 500 Wavelength, nm

Figure 60. The time dependent spectra resulting from irradiation of a MPD-henzhydryl bromide-1,2-DCE solution under vacuum (#,x,f) and under oxygen (o). The concentration of MPD is 0 .0 1 6 5 M. A 80 ns pulse was used. 151

22 20 t=0 .3 V-S, under vacuum

^1^14

'difference \spectrum ^

250 '0 280 290 300 310 320 330 340 350 Wavelength, nm

Figure 6I. The optical absorption spectra resulting from irradiation of a 0.27 M solution of MPD in DOE under vacuum and a 0.20 M solution of MPD in DOE in the presence of 0.002 M of ammonia. A difference spectrum (0 , was obtained by normalizing the two spectra at 271 nm and then taking their difference. 80 ns pulses were used. 152

the absorption spectrum of this species as shown in Figure 6l.

It has an absorption peak at about 302-307 nm similar to that

of tetrasubstituted allyl cation generated by irradiation of

DMPD-DCE solution. This species, as well as the species with

an absorption peak at 312 nm in Figure 60, should be the

trisubstituted allyl cation generated in reaction (6 5) and

reaction (66) respectively.

Attempt has been tried to measure the formation rate of

the product ion of reaction (64). Unfortunately, this ion has a much faster decay rate than tetrasubstituted allyl cation has, which, together with the complication due to overlapping absorption, excludes the possibility of measuring its formation rate.

Although the evidences are not as conclusive as in the

case of DMPD, on the basis of their similarities, and the information presented above, we tentatively assigned the transient species with an absorption peak at 312 nm in Figure

60 as the product ion of reaction (64), a trisubstituted allyl cation.

(2) Reactivity.

The rate constants between benzhydryl cation and methyl substituted 1 ,3-conjugated dienes were measured by the similar procedure described before. The plots of pseudo-first order rate constants versus concentration of 1,3-conjugated dienes 153 are shown in Figure 62 and Figure 6 3. The second order rate constants are listed in Table 8.

The rate constant between benzhydryl cation and 1,3-buta diene was estimated to be smaller than 10^ M“^sec“^ based on the fact that the decay rate of benzhydryl cation did not change significantly when the concentration of 1 ,3-butadiene was increased up to 0.52 M.

The rate constant between benzhydryl cation and DMPD was measured by monitoring both the decay of benzhydryl cation at # 9 nm and the formation of tetrasubstituted allyl cation at 312 nm. As shown in Figure 59» the absorption of this allyl cation was overlapped by the absorption of benzhydryl radical. Since benzhydryl radical was scavenged by oxygen while allyl cation was not, they were separated by irradiating oxygen saturated 1,2-DCE solution. The formation curve of allyl cation at 312 nm was obtained by substracting the decay curve of benzhydryl radical at 312 nm, point by point, from the total rate curve at 312 nm. The decay curve of benzhydryl cation was obtained by irradiating a benzhydryl bromide-DCE solution saturated with oxygen in the absence of

DMPD. It was assumed that the addition of DMPD did not change the decay rate of benzhydryl radical. An example of this pointwise substraction is shown in Figure 64.

The pseudo-first order rate constant of the formation process was obtained by plotting In CD versus time from the formation curve of allyl cation obtained above. Two Ilsoprene] or [DMBD] x 10^ M 2 3 10 11 12 13 14 15 16 17

12 s 10

Figure 62. The plots of pseudo-first order rate constants versus concentrations of isoprene(o, DMBD(a , -- ), and MPD(#, ----- ) for the reactions between benzhydryl cation and isoprene, DMBD, and MPD in 1,2-DCE at 24 0. W 12

X M

Figure 6 3. The plot of pseudo-first order rate constant versus concentration of DMPD for the reaction between benzhydryl cation and DMPD in 1,2-DCE at 24 G . Solid circles (•) were obtained by monitoring the decay of benzhydryl cation at 449 nm. Crosses (X) were obtained by monitoring the formation of tetrasubstituted allyl cation at 312 nm. Table 8. Rate constants for reactions between benzhydryl cation and 1^3-conjugated dienes in 1,2-DCE at 24 C.

BD MBD DMBD MPD DMPD GH. GH_ CH2=CH-CH=CHg CH2=CH-G=CH2 CH2=C-(j!=GH2 -:^)G=GH-GH=GHp ^'"G=GH-G=GHp , CH^ H^G GH^ G H ^ ^

PhgCH® < 10^ 7.1 X 10^ 2.7 X lo"^ 2.5 X 10® 1.0 X 10^ i±15fo) {±15%) (±15%) (±15%)

CHq product ion CHgR GHgR GH^' GH^ ‘gH^R

R represents PhpCH. 157 0.1

0 1 2 3 Time, us

Figure 64. The formation rate curve of tetrasubstituted allyl cation at 312 nm (•) was obtained by substracting the decay curve of benzhydryl radical under oxygen (x) from the composite rate curve monitored at 312 nm (o). 158

pseudo-first order formation rate constants were obtained

at lower DMPD concentration and were plotted against concen

tration of DMPD as shown in Figure 6 3. At higher concentra

tion of DMPD it is difficult to get accurate data because of

two reasons ; (a) The method described above to separate the

contribution of benzhydryl radical from that of allyl cation

introduces a larger uncertainty when the composite rate curve

looks "flat" (i.e. at higher concentration of DMPD the forma

tion rate becomes so fast that it cancels with the decay rate of benzhydryl radical), (b) At very high concentration

of DMPD, the assumption that the decay rate of benzhydryl radical is not changed by the addition of DMPD may not be valid.

For the above reasons, only two points, representing a two-fold change of concentration, were obtained. As shown in Figure 6 3, the formation rate of tetrasubstituted allyl cation is approximately equal to the decay rate of benzhydryl cation in the limited concentration range that examined.

The second order rate constants listed in Table 8 range from

It is most effective to stabilize the allylic cation by substituting methyl group at the end carbon of allylic system.

This is evident by the rate difference, one order of magnitude, between DMBD and MPD.

Finally, all the rate constants that have been discussed above are summarized in Table 10. Table 10, Summary of the rate constants for the reactions between arylcarbénium ions and alkenes in 1,2-DCE at 24 C. k X 10^, M“^sec“^

l l T ' hexeê benêê O O O ' ’^' DMBD MPD DMPD

PhCHg <10^ 1.9 19 2.2 1.5 9.4 27 --- 0.8? — — — —

PhgCH® — <10^ 9.5 — - 15 74 <10^ 7.1 27 250 1000

All the rate constants have an uncertainty about +l^fo. The rate constant between benzyl cation and BD and the rate constant between benzhydryl cation and 1 ,3-cyclohexadiene have an uncertainty about +25^. APPENDIX A

COMPETITION KINETICS — TWO SOLUTES COMPETING FOR ONE SOLVENT

CATION

The reaction scheme is as follows :

^ + P > P® [p]3>[s+]

s'*’ + R — > product

where S'*’ represents solvent cation, P represents the precursor of arylcarbénium ion, R represents the second solute added as competition reagent, and P® represents arylcartenium ion.

In the absence of R:

— = - kj_[s'^][pJ = - k£[s'*’J , where = k^[p]

[S"]= [S+]^ e-^i^

[P®] - i S + ] _ [ S + ] o

= [S+]p(l-e-y)

In the presence of R:

' ~ ] - k^IS ] , where k^ = k ^ ^ j

[S+] . [S+Jo e-(ki + k p t

161 l62

kA)t . k'[S+]^

[p®] = - M f l L (i-e-'ki+kp %i+k2

Using *'s to indicate concentrations and optical densities in the presence of R

[P®J ^ (i_e-(k;+%2)t)

As t becomes large, this equation simplifies to

[p®]„_ k'+k^

where [P® j^represents the concentration of P® at time =00 (i the maximum concentration of P® that may be yielded) in the absence of R. The above equation leads to

> , . kg [R]

0°;® k^ [p]

By plotting (OD ^ /0D*^)-1 versus [Rj/[Pj , a straight line should be obtained with slope equal to k^/k^. 163

APPENDIX B

COMPETITION KINETICS-TWO SOLUTES COMPETING FOR TWO SOLVENT

CATIONS

The reaction scheme to be considered is;

P ■

Sj + P ^ P k_ + R --^ product

+ kh Sp + R > product'

where S^ and represent two different solvent cations, P represents the precursor of arylcarbénium ion, R represents the competition reagent, and P® represents arylcarbenium ion.

In the absence of R;

d[stl = - k|[S^] , where = kj^fpJ dt

= - k^[Sg] , where = kg[p]

lstJ= [Sp,

ls;i= e-kgt

■ [P®] = [S[l„ ^ [S^]„ - [sp - [sp 164

As t becomes large

In the presence of R:

= -(kJ+k^)[Sp , where k^=k^[Pl

= -(k2+kj^)[sp , where k^=k^^;R]

; IS*]= [S^], e-(k^k^)t

[s;i= [s;], e-(ki+k^t

^ * k^Is;! = k-rs^]^ e-(ki^k^)t + k^rg+]^

[P®]= J i Ü l k . (i-e-(k^k')t)^ ’'^•^2-° (l.e-":2+kpt) k^+k^ k|+k4

As t becomes large

Using *'s to indicate concentrations and optical densities in the presence of R

L ? % . [s;], + [s;]. 165

° V y(k^[Rj/iPj) -®l-o * k2+(kjR]/fPJ) '4îo

In order to further simplify this equation, the relative 166

APPENDIX C

PSEUDO-FIRST ORDER RATE EQUATION WITH OVERLAPPING ABSORPTION

[R3»[P®j

■where P represents arylcarbenium ion and R represents reactant.

= -k[P®][R] = -k’[P®], where k'=k[R]

InfP®] - ln[P^]^ = -k't

In OD , - ln(OD = -k't p* p® o

Let OD=OD^+ OD^ , where OD = total optical density at the •wavelength monitored.

OD ^ = optical density that only

due to P®.

OD = overlapping absorption

In(OD-OD^) - In(OD-OD^)^ = -k't

If overlapping absorption is independent of time (i.e.

OD^ =(0D^)o = OD^ ) then In(OD-OD^) - ln(OD^-0%) = -k't

By plotting -ln(OD-OI^) versus time, a straight line should be obtained with slope=k' and intercept=ln(OD -0D»). 167

APPE^roiX D

CALCULATIONS O F A H ’s FOR REACTIONS BETWEEN BENZYL CATION

AND ALKENES

The overall reaction

-C-C-CHgPh A H

may he broken down into the following reactions;

-4 PhCHo^2 ^^1

-C-C- AH,

PhCHg' + -C-C-----> -C-C-CHgPh AH^

The calculations of AH's are shown in Table 9. Table 9* Calculations of AH's for 168 reactions between benzyl cation and alkenes in gas phase.

PhCH® + -C=C- -C-C-CHgPh, AH(Kcal/mole)

AH, AH 2 AH^ AH^ AH ethylene 5 9.3^ 200.4^ +1 2 .5 propylene 5 9.1^ 182.2^ -5 .2 . 182.9^ -5 .9 -6 i .5^ isobutylene -17 8.7^ 5 7.9^ 1 7 1.8^ -1 7 .5 171.1 -18.2 cyclohexene 58.4^ 1 7 6.7® -12.1

1,3-butadiene 5 0.4^ -6 3.5''

a. This value is the average of 178.3 (from reference 101) and 179*1 (from reference ll4). b. from reference ll4.

c. C-C bond energies for CH^CH^-CH^Ph, CH^CH CH2"CH2Ph, (CH^)2ÔCH2-CH2Ph, and C^H^Q-CH2Ph were estimated to be -6 8 .5 Kcal/mole which is the average value of the bond energies of C2H^-CH2Ph(6 8.7), n-C^H^-CH2Ph(69), and i-C^Hy-CH2Ph(6 7*8) (from reference 114). The bond energy of CH2CHCHCH2-CH2Ph was estimated to be 6 5 .5 Kcal/mole.

These values were estimated from ionization potentials of Cÿly -200.4, (CH^)2CH*-182.2, (C2H^)(CH^)CH.-182.9, “ CH2^CH-CH,.-188.2, -- -- Kcal/mole (from reference 72,113).

from reference 101. REFERENCES AND NOTES

1. L. M. Dorfman, Techniques of Chemistry, Vol. 6, Part 2, A. Weissberger and G. Haïmes, eds., John Wiley & Sons, New York, 1973i Chapter 11.

2. M. S. Matheson and L. M. Dorfman, Puise Radiolysis, MIT Press, Cambridge, 1 9 6 9.

3 . j. Boag, Trans. Faraday Soc., 677(1967).

4. Operation Manual for Varian V-7715A electron linear accelerator.

5. B. Gall, Ph.D. Dissertation, The Ohio State University, 1968.

6. R. P. Bell, J. Chem. Soc., 1371(1931).

7. Y. Wang, Master Thesis, The Ohio State University,. 1976.

8. L. M. Dorfman and I. A. Tauh, J. Amer. Chem. Soc., 8 5, 2370(1963).

9 . N. Shank, Ph.D. Dissertation, The Ohio State University, 1 9 6 9.

10. A. Hantzsch, Chem. Ber., j4, 2573(1921).

11. M. Gomherg and L. H. Cone, Ann. Chem., 370, l42, 193(1909); 376, 183(1910).

12. A. Baeyer and V. Villiger, Chem. Ber., 3^, II8 9, 3013(1902). 169 170

13- C. D. Nenitzescu in Carbonium Ions, Vol. I, G. A. Olah and P. V. R. Schleyer, eds., John Wiley & Sons, 1968, Chapter 1.

14. D. Bethell and V. Gold, Cairbonium Ions, An Introduction, Academic Press, 1967-

15- G. A. Olah and J. A. Olah in Carbonium Ions, Vol. II G. A. Olah and P.V.R. Schleyer, eds., John Wiley & Sons, 1 9 6 8, Chapter 17.

1 6. For the methods of investigation of carbonium ions, see Carbonium Ions, Vol. I, G. A. Olah and P. V. R. Schleyer, , eds., John Wiley & Sons, 1 9 6 8.

1 7. Carbonium Ions, Vol. I-V, G. A. Olah and P. V. R. Schleyer, eds., John Wiley & Sons, New York, I9 6 8.

1 8. R. A. Diffenbach, K. Sano, and R. W. Taft, J. Amer. Chem. Soc., 8 8, 4 7 4 7(1966).

1 9. For complete references see reference 7.

20. C. Swain, C. Scott, and K. Lohman, J. Amer. Chem. Soc., 7i. 136(1953).

21. R. L. Jones and L. M. Dorfman, J. Amer. Chem. Soc., 9 6, 57 1 5(1974).

22. R. J. Sujdak, R. L. Jones, and L. M. Dorfman, J. Amer. Chem. Soc., ^8, 4875(1976).

2 3. L. M. Dorfman, Y. Wang, H. Y. Wang, and R. J. Sujdak, Faraday Discussion No. 6 3, Radiation Effects-63/l3(l976). 171

24. L. M. Dorfman, V. M. DePalma, and Y. Wang, in Protons and Ions Involved in Fast Dynamic Phenomena, Elsevier Scientific Publishing Co., 1978, Proceedings of 30th Meeting of Société de Chimie Physique, Paris, November 1977.

2 5. V. M. DePalma, Y. Wang, and L. M. Dorfman, J. Amer. Chem. Soc., 100, 5416(1978).

2 6. J. W. T. Spinks and R. J. Woods, An Introduction to Radiation Chemistry. John Wiley & Sons, New York, 1976.

2 7. P. Ausloos, eds., Fundamental Processes in Radiation Chemistry. John Wiley & Sons, New York, 1 9 6 8.

28. T. Shida, J. Phys. Chem., 72, 2 59 7(1 968).

2 9. S. Arai and L. M. Dorfman, J. Chem. Phys., 2190(1964).

3 0 . J. Brandon and L. M. Dorfman, J. Chem. Phys., 3849 (1970).

3 1. W. H. Hamill in Radical Ions, E. Kaiser and L. Kevan, eds., Interscience, New York, 1 9 6 8, Chapter 9 .

3 2. T. Shida and W. H. Hamill, J. Chem. Phy., # , 2375(1966).

33. S. Arai, A. Kira, and M. Imamura, J. Phys. Chem., 80.1968 (1976).

3 4. A. Kira, T. Nakamura, and M. Imamura, J.Phys. Chem., 8l, 5 9 1(1977).

3 5. S. Arai, H. Ueda, R. Firestone, and L. M. Dorfman, J. Chem. Phys., iO, 1072(1969). 172

36. N. Shank and L. M. Dorfman, J. Chem. Phys., j2, 4441(1970).

37. T. Shida and W. H. Hamill, J. Chem. Phys., # , 4372(1966) .

38. T. Shida and S. Iwata, J. Amer. Chem. Soc., 3473(1973).

39. (a) A. Kira, T. Nakamura, and M. Imamura, J. Physl Chem. 82 , 1961(1978). (h) S. Arai, M. Imamura, submitted to J. Phys. Chem.

40. G. A. Olah, C. U. Pittman, Jr., and M- C. R. Symons in Carbonium Ions, Vol. I, G. A. Olah and P. V. R. Schleyer, eds., John Wiley & Sons, 1 9 6 8, P. l64.

41. P . 157 of reference 40.

42. R. J. Sujdak, Ph.D. Dissertation, The Ohio State University 1 9 7 7.

4 3. G. A. Olah, Halonium Ions, John Wiley & Sons, New York, 1 9 7 5.

44. S. D. Beatty, S. D. Worley, and S. P. McManus, J. Amer. Chem. Soc., 100, 4254(1978).

4 5. W. J. Hehre and P. C. Hiberty, J. Amer. Chem. Soc., 9 6. 2665(1974).

46. C. C. Van de Sande, F. W. McLafferty, J. Amer. Chem. Soc., iz, 2298(1975).

4 7. D. W. Berman, V. Anicich, and J. C. Beauchamp, submitted to J. Amer. Chem. Soc.

48. p. 197 of reference 40. 173

49. The ionization potential of styrene is 8.86 ev (from p.158 of reference 101); the ionization potential of 1,1-diphenylethylene is expected to be smaller than.that of styrene.

50. P. 159-P* 165 of reference 40.

5 1. P. 168 of reference 101.

5 2. T. Shida and W. H. Hamill, J. Amer. Chem. Soc., 5371, 5 3 7 6(1966).

5 3. N. C. Deno, J. Bollinger, N. Friedman, K. Hafer, J. D. Hodge, and J. J. Houser, J. Amer. Chem. Soc., 8^, 2998 (1963).

5 4. K. C. Kim, J. H. Beynon, and R. G. Cooks, J. Chem. Phys., 6 1, 1 305(1974).

5 5. Z. Luczynski and H. Wineel, Int. J. Mass. Spectrom. Ion Phys., 14,29(1974).

5 6. F. W. McLafferty, Anal. Chem., 34, 2(1962).

5 7. P. 113 of reference 43.

5 8. Z. Luczynski and J. A. Herman,J. Phys. Chem., 1679 (1978).

5 9. T. 0. Tiernan and B. M. Hughes, Advan. Chem. Ser., 82, 412 (1968).

6 0. J. L. Beauchamp, D. Holtz, S. D. Woodgate and S. L. Patt, J. Amer. Chem. Soc., 9^ , 2798(1972). 174

61. H. Lischka and H-J Kohler, J. Amer. Chem. Soc., 5297 (1978).

62. A. C. Hopkinson, M. H. Lien, K. Yates, and I. G. Csizmadia, Theor. Chim. Acta., 21(1975)*

6 3. A. C. Hopkinson, M. H. Lien, K. Yates, and I. G. Csizmadia, Theor. Chim. Acta., 385(1977)*

64. A. Ledwith and D. C. Sherrington, Kinetics of Homogeneous Cationic Polymerization, in Comprehensive Chemical Kinetics, C. H. Bamford and C. F. H. Tipper, eds., Elsevier Scientific Publishing Co., 1976, p . 67 -p.127.

6 5. A. Ledwith and D.C. Sherrington, in Reactivity. Mechanism and Structures in Polymer Chemistry, A. D. Jenkins and A. Ledwith, eds., John Wiley & Sons, London, 1974, p.244.

66. P. H. Plesch, ed., The Chemistry of Cationic Polymeriza tion, Pergamon Press Ltd., London, 1 9 6 3.

6 7. P. H. Plesch, ed., Cationic Polymerization and Related Complexes. W. Heffer & Sons, Cambridge, 1953*

68. B. L. Erusalimskii, Modern Views on the Mechanism of the Cationic Polymerization of Vinyl Monomers. Rues. Chem. Rev., 651(1963).

6 9. J. D. Kennedy and A. W. Langer, Recent Advances in Cationic Polymerization, Advan. Polymer. Sci., 2* 508 (1964).

7 0. A. M. Eastham, in Encyclopedia of Polymer Science and Technology, Vol. 3, Wiley-Interscience, New York, 1 9 6 5, p.3 5. 175

71. J. E. Wilson, Radiation Chemistry of Monomers, Polymers, and Plastics, Marcel Dekker, Inc., New York, 197^.

72. Hideo Sawada, Thermodynamics of PolT/merization, Marcel. Dekker, Inc., New York, 1976.

73. A. Chapiro, Radiation Chemistry of Polymeric Systems, in High Polyners, Vol. XV, Wiley-Interscience. New York, 1962.

74. J. P. Kennedy, Cationic Polymerization of Olefins; A Critical Inventory. John Wiley & Sons, New York, 1975-

75. A. Ledwith and D.C. Sherrington, Adv. Polymer. Sci., 19. 1(1974).

76. D.C. Pepper, R. E. Burton, M. J. Hayes, A. Albert, and D. H. Jenkinson, Proc. Roy. Soc. of London Ser. A, 263, 58, 63,75,82(1961).

77. D. C. Pepper and P. J. Reilly, Proc. Roy. Soc. of London Ser. A, 2gl, 41(1966).

78. T. Masuda, M. Sawamoto, and T. Higashimura, Makromol. Chem., m , 2981, 2995(1976).

79. M. Sawamoto, T. Masuda, T. Higashimura, S. Kobayashi, and T. Saegusa, Makromol. Chem., 178. 389(1977).

80. N. Kanoh, T. Higashimura, and S. Okamura, Makromol. Chem., i6, 65(1962).

81. F. S. Dainton, in Cationic Polymerization and Related Complexes. P. H. Plesch, ed., Heffer & Sons, Cambridge, 1953, p.148. 176

82. P. H. Plesch, M. Polanyi, and H. A. Skinner, J. Chem. Soc., 25 7(1947).

8 3. A. G. Evans and G. W. Meadows, Trans. Faraday Soc., 46, 327(1950).

84. A. Ledwith, Advan. Chem. Ser., 21> 317(1969).

8 5. C. E. H. Pawn, C. Fitzsimmons, A. Ledwith, J. Penfold, D.C. Sherrington, and J. A. Weightman. Polymer, 1^, 119

. (1971).

86. A. Ledwith, Symposium Lecture Honoring Michael Szwarc, May, 1 9 7 8, Pure and Applied Chemistry.

8 7. F. W. Lampe, J. Phys. Chem., 6 3, 1986(1959).

88. W. H. T. Davison, S. H. Pinner, and R. Worrall, Chem. & Ind., 1274(1957).

8 9. E. Collinson, F. S. Dainton, and H. A. Gillis, J. Phys. Chem., 90 9 (1959).

9 0 . K. Hayashi, K. Hayashi, and S. Okamura, J. Polym. Sci. Part A-1, Vol. 9, 2305(1971).

9 1. M. A. Bonin, W. R. Busier, and Ff. Williams, J. Amer. Chem. Soc., 87 . 199(1965).

9 2. E. Hubmann, R. B. Taylor, and Ff. Williams, Trans. Faraday Soc., 88(1966) .

9 3 . Ff. Williams, K. Hayashi, K. Ueno, K. Hayashi, Y. Yamazawa, T. Takagaki, and S. Diamura, Trans. Faraday Soc., 6 3. 1478, 1489, 1501(1967). 177

94. R. B. Taylor and Ff. Williams, J. Amer. Chem. Soc., 91. 3728(1969).

95. M. Sawamoto and T. Higashimura, submitted to macromolecules.

9 6. E. J. Burrell, J. Phys. Chem., 68, 3885(1964).

9 7. J. P. Keene, E. J. Land, and A. J. Swallow, J. Amer. Chem. Soc., 87, 5284(1965).

9 8. A. J. Swallow, in Radiation Chemistry-II, Advan. Chem. Ser., 82, 499(1968).

9 9 . C. Schneider and A. J. Swallow, Polymer Letters, 4, 277 (1966).

100. S. Egusa, S. Arai, A. Kira, M. Imamura, and Y. Tabata, Rad. Phys. & Chem., 419(1977).

101. V. I. Vedeneyev, L. V. Gurvich, V. N. Kondratyev, V. A. Medvedev, and Y. L. Frankevich, Bond Energies, Ionization Energies, and Electron Affinities, translated by Scripta Technica Ltd., 1 9 6 2.

102. p. 187 of reference 40.

1 0 3. p. 179 of reference 40.

104. A. G. Evans and M. Polanyi, J. Chem. Soc., 252(1947).

1 0 5. M. L. Burstall, in The Chemistry of Cationic Polymerization. P. H. Plesch, ed., Pergamon Press Ltd., London, 1 9 6 3, p.79*

1 0 6. J. W. Baker and W. S. Nathan, J. Chem. Soc., 1844(1935). 178

107. J. W. Baker, Hypercon.jugation, Oxford University Press, 1952.

108. J. W. Larsen, P. A. Bonis, M. W. Grant, C. A. Lane, J. Amer. Chem. Soc., 2067(1971).

1 0 9. L. Radom, J. A. Pople, V. Buss, and P. V. R. Schleyer, J. Amer. Chem. Soc., 92, 6380(1970).

110. L. Radom, J. A. Pople, V. Buss, and P. V. R. Schleyer, J. Amer. Chem. Soc., 6987(1970).

111. L. Radom, J. A. Pople, and P. V. R. Schleyer, J. Amer. Chem. Soc., 94, 5935(1972).

112. L. Radom, Aust. J. Chem., 231(1974).

1 1 3. D. Bethell and V. Gold, Carhonium Ions, Academic, New York, 1 9 6 7, p. 6 7'.

114. K. W. Egger and A. T. Cocks, Helv. Chim. Acta., 5 6, 15 1 6, 1537(1 973).

1 1 5. J. Hine, "Structural Effects on Equilibria in Organic Chemistry", Wiley-Interscience, New York, 1975.

1 1 6. L. P. Hammett, "Physical Organic Chemistry, Reaction Rates, Equilibria, and Mechanisms", McGraw-Hill, New York, 1 9 7 0.

1 1 7. J. Hine, "The principle of Least Nuclear Motion", Advan. Phys. Org. Chem., 1(1977).

1 1 8. F. 0. Rice and E. Teller, J. Chem. Phys., 6, 489(1938).