Infrared multiple-photon dissociation of small organic sulfoxides

(1)

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of th is reproduction is d e p en d en t upon the quality of th e copy subm itted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6” x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

Bell & Howell Information and Learning

3(X) North Zeeb Road, Ann Artx)r, Ml 48106-1346 USA

(2)

(3)

by

Marko Peter Banjavcic

B .M ath., University of W aterloo, 1988 M.Sc., University of W aterloo, 1990

A Thesis S ubm itted in P artial Fulfillment of the Requirem ents for the Degree of

D O C T O R O F P H IL O S O P H Y

in the

D epartm ent of Chem istry We accept this thesis as conforming

to the required standard

Dr. T.^E^-Gough, Si^tervisofcj (D epartm ent of C hem istry)

Dr. W. J. Balfour, D epartm ental M em ber (D epartm ent of C hem istry)

Dr. C. Qian, D epartm ental M ember (D epartm ent of C hem istry)

Dr. G.A. Beer, O utside M em ber (D epartm ent of Physics)

Dr. S.V. Filseth, E xternal Exam iner (D epartm ent of Chem istry, York University)

(4)

Supervisor: Dr. T .E . Gough

A b str a c t

T he infrared m ultiple-photon dissociation of several sm all organic sulfoxides was stu d ied in order to provide some insight in th e dissociative product channels for this class of molecules. V ibrational excitation was achieved w ith a TEA C O ; laser and photofrag m ents were identified w ith a laser ionized tim e of flight mass spectrom eter. A beam of 10.5 eV photons generated in a static xenon gas tripling cell from a tightly focused NdiYAG th ird harm onic beam of light was used to ionize the molecules and fragm ent species.

T he photoionization mass sp ectra were found to contain fewer fragm entation species th an the corresponding electron im pact ionization m ass spectra. Numerous product species were observed from skeletal rearrangem ent reactions despite the lower level of energy excitation from th e photoionization process.

T h e infrared photolysis wavelength dependence of d im eth y l sulfoxide indicated th a t the m axim um abundance for th e m ajo r fragm entation species occurred a t (or near) 1085.8 cm “ ‘. T h e m ajor products a t 1085.8 cm “ ^ were [HaCSO]"*" and [CHa]'*'. M inor products were also observed a t a higher o u tp u t power threshold like th e skeletal rearrangem ent species [HaCS]'*' and [OCHa]'*'.

W ith the infrared photolysis wavelength dependence of m eth y l phenyl sulfoxide, th e m ajo r product species ([CHa]'*' and [OSCHa]'*') had a m axim um abundance near

(5)

majcimum abundance near 1080 cm ~ ‘ while large mass fragm ent species ([CgHgS]'^* and [CgHsSO]'"') had a m axim um abundance th a t was fu rth er red-shifted (1058.9 and 1050.4 c m " \ respectively). The photolysis a t 1085.8 cm “ ‘ generated th e high mass

fragm ent species ([CgHsSOj^, [CgHgS]^* and [CsHsS]'*’) a t all CO3 laser o u tp u t pow

ers bu t a p lateau in th e abundance was observed a t higher o u tp u t powers. T he sm aller mass fragm entation species only appeared after a threshold power was surpassed.

T he m axim um abundance for th e various fragm entation species of sec-butyl m ethyl sulfoxide occurred near 1072 cm ~^ T he C O ; laser o u tp u t power dependence for the form ation of th e butane, butyl and butene ions indicated all three were generated at all o u tp u t powers w ith an abundance plateau or decline occurring a t higher o u tp u t powers. Fragm entation species were observed from the butyl group prim ary products w ith th e power threshold increasing for th e sequentially sm aller secondary species.

(6)

Exam iners:

Dr. T. E. G ougbvSupervisor (D epartm ent of C hem istry)

Dr. W. J. Balfour, D epartm ental M ember (D epartm ent of C hem istry)

Dr. C. Q ian, D epartm ental M ember (D epartm ent of C hem istry)

Dr. G .A . Beer, O utside M ember (D epartm ent of Physics)

__________________________________________________ Dr. S.V. Filseth, E xternal Exam iner (D epartm ent of Chem istry, York University)

(7)

A bstract

ii

Table o f C ontents

v

List of Tables

viii

List of Figures

ix

Acknow ledgem ents

xv

D edication

xvii

1 Introduction

1

1.1 Mass S p e c tro m e te rs... 4

1.1.1 In s tru m e n ta tio n ... 4

1.1.2 Uni molecular R eaction T h e o ry ... 10

1.2 Laser E x c ita tio n ... 21

(8)

1.2.2 Infrared M ultiple-Photon Dissociation ... 27

1.2.3 Laser In itiated Uni m olecular D isso ciatio n ... 31

1.3 Sm all O rganic Sulfoxide M olecules... 35

1.3.1 Electron Im pact Mass S p e c tro m e try ... 36

1.3.2 Sulfoxide Photolysis ... 39

2 Experim ental Apparatus

45

2.1 T im e of Flight Vacuum C h a m b e r ... 48

2.2 Photoionization of M o lecu les... 50

2.3 T im e of Flight Mass S p e c tro m e te r... 53

2.4 V ibrational Excitation of M o le c u le s ... 56

2.5 Mass S p ectra P o s t-P ro c e s s in g ... 65

3 E xperim ental R esults

68

3.1 Electron Im pact Ionization and Photoionization Maiss S pectra . . . . 69

3.1.1 D im ethyl S u lf o x id e ... 70

3.1.2 D iethyl S u lf o x id e ... 73

3.1.3 M ethyl Phenyl S u lf o x id e ... 75

3.1.4 E th y l Phenyl S u lfo x id e ... 77

3.1.5 fert-B utyl M ethyl S u l f o x i d e ... 79

3.1.6 sec-Butyl M ethyl S u lfo x id e ... 81

3.1.7 sec-Butyl E thyl S u lfo x ide... 84

(9)

3.2 D im ethyl Sulfoxide Infrared P h o to l y s is ... 89

3.2.1 W avelength D e p e n d e n c e ... 90

3.2.2 Power D e p e n d e n c e ... 104

3.3 M ethyl Phenyl Sulfoxide Infrared P h o t o l y s i s ... 116

3.4 sec-Butyl M ethyl Sulfoxide Infrared P h o to ly s is ... 145

4 C oncluding Remarks

175

4.1 Mass S p e c t r a ... 176

4.2 D im ethyl Sulfoxide P h o to ly s is ... 179

4.3 M ethyl Phenyl Sulfoxide P h o to ly s is ... 181

4.4 sec-Butyl M ethyl Sulfoxide P h o to ly s is ... 183

4.5 Final Com m ents ... 185

(10)

List o f Tables

3.1 Fragm ent Sum m ary for Species G enerated from D im ethyl Sulfoxide . 91

3.2 Fragm ent Sum m ary for Species from G enerated M ethyl Phenyl Sulfoxidell? 3.3 Fragm ent Sum m ary for Species from G enerated sec-Butyl M ethyl Sulf

(11)

List o f Figures

2.1 E xperim ental A pparatus - Top V i e w ... 46

2.2 E xperim ental A pparatus - Side V i e w ... 47

2.3 V ibrational Modes for the C O ; M o le c u le ... 58

2.4 T he M ajor Transitions for th e C O ; L a s e r ... 60

3.1 D im ethyl Sulfoxide Photoionization and Electron Im p act Mass S pectra 71 3.2 Diethyl Sulfoxide Photoionization and Electron Im p act Mass S p ectra 74 3.3 M ethyl Phenyl Sulfoxide Photoionization and E lectron Im pact Mass Spectra ... 76

3.4 E thyl Phenyl Sulfoxide Photoionization and E lectron Im pact Mass Spectra ... 78

3.5 (ert-B utyl M ethyl Sulfoxide Photoionization and E lectron Im pact Mass Spectra ... 80

3.6 sec-Butyl M ethyl Sulfoxide Photoionization and E lectron Im pact Mass S pectra ... 83

(12)

3.7 sec-B utyl E th y l Sulfoxide Photoionizatioa and Electron Im pact Mass

S p e c tra ... 85

3.8 D im ethyl Sulfoxide T im e of Flight Mciss Spectra (9R32 C O ; Laser Line) 93

3.9 D im ethyl Sulfoxide Mass S p ectra (9R32 C O ; Laser L i n e ) ... 94

3.10 D im ethyl Sulfoxide T im e of Flight Mass S pectra (9P I6 C O ; Laser Line) 97

3.11 D im ethyl Sulfoxide Mass S p ectra (9P16 C O ; Laser L i n e ) ... 98

3.12 D im ethyl Sulfoxide C O ; Leiser W avelength D ependent Photolysis for

Masses 63, 15 and 48 a m u ... 99

3.13 D im ethyl Sulfoxide C O ; Laser W avelength D ependent Photolysis for

Mass 48 am u ... 101

3.14 D im ethyl Sulfoxide C O ; Laser W avelength D ependent Photolysis for

M ass 45 am u ... 102

3.15 D im ethyl Sulfoxide C O ; Lciser W avelength D ependent Photolysis for

M asses 31 and 47 a m u ... 103

3.16 D im ethyl Sulfoxide Tim e of Flight Meiss S pectra (9R32 C O ; Laser

Line, 1 . 6 W ) ... 106

3.17 D im ethyl Sulfoxide Mass S pectra (9R32 C O ; Laser Line, 1.6 W ) . . . 107 3.18 D im ethyl Sulfoxide Tim e of Flight Mass Spectra (9R32 C O ; Laser

Line, 0.1 W ) ... 109

(13)

3.20 D im ethyl Sulfoxide C O ; Laser O u tp u t Power D ependent Photolysis for

Masses 63 and 15 a m u ... 112

Mass 48 am u ... 113

Masses 31 and 47 a m u ... 114

Mziss 45 am u ... 115

3.24 M ethyl Phenyl Sulfoxide T im e of Flight Mass S pectra (9R32 C O ; Laser

L i n e ) ... 119

3.25 M ethyl Phenyl Sulfoxide Mtiss S pectra (9R32 C O ; Laser Line) . . . . 121

3.26 M ethyl Phenyl Sulfoxide Tim e of Flight Mêiss S pectra (9P22 C O ; Laser

L i n e ) ... 124

3.27 M ethyl Phenyl Sulfoxide Mass Spectra (9P22 C O ; Laser Line) . . . . 125

3.28 M ethyl Phenyl Sulfoxide C O ; Laser W avelength D ependent Photolysis

for Meiss 63 am u ... 127

for Mass 15 a m u ... 128

(14)

3.31 M ethyl Phenyl Sulfoxide C O j Laser W avelength D ependent Photolysis

for Masses 45 and 53 a m u ... 131

3.32 M ethyl Phenyl Sulfoxide C O j Laser Wavelength D ependent Photolysis

for Mass 112 a m u ... 133

3.33 M ethyl Phenyl Sulfoxide Tim e of Flight Mass Spectra (9R32 C0% Lciser

Line, 1 . 6 W ) ... 136

3.34 M ethyl Phenyl Sulfoxide Mass Spectra (9R32 CO2 Laser Line, 1.6 W) 137

3.35 M ethyl Phenyl Sulfoxide Tim e of Flight Mass S pectra (9R32 C O ; Laser

Line, 0.1 W ) ... 138

3.36 M ethyl Phenyl Sulfoxide Mass S pectra (9R32 C O ; Laser Line, 0.1 W) 139 3.37 M ethyl Phenyl Sulfoxide C O ; Laser O u tp u t Power D ependent Photol

ysis for Masses 63 and 77 a m u 141

3.38 M ethyl Phenyl Sulfoxide C O ; Laser O u tp u t Power D ependent Photol

ysis for Masses 45, 53 and 51 a m u 142

3.39 M ethyl Phenyl Sulfoxide C O ; Laser O utp u t Power D ependent Photol

ysis for Mcisses 15 and 125 a m u ... 144

3.40 M ethyl Phenyl Sulfoxide C O ; Laser O utp u t Power D ependent Photol

ysis for Mass 97 a m u ... 146

3.41 M ethyl Phenyl Sulfoxide C O ; Laser O u tp u t Power D ependent Photol

(15)

3.42 sec-Butyl M ethyl Sulfoxide T im e of Flight Mass S pectra (9R32 C O j

Lziser Line) ... 151

3.43 sec-Butyl M ethyl Sulfoxide Mass S pectra (9R32 C O2 Laser Line) . . . 152

3.44 sec-Butyl M ethyl Sulfoxide T im e of Flight Mass S pectra (9P34 C O ;

Laser Line) ... 153

3.45 sec-Butyl M ethyl Sulfoxide Mass S pectra (9P34 C O ; Laser Line) . . . 154

3.46 sec-Butyl M ethyl Sulfoxide C O ; Laser W avelength D ependent Photol

ysis for Masses 57 and 63 a m u ... 155

ysis for Mcisses 15 and 29 a m u ... 160

ysis for Mzisses 41 and 42 a m u ... 161

3.52 sec-Butyl M ethyl Sulfoxide T im e of Flight Mass S p ectra (9R32 C O ;

Laser Line, 1.6 W ) ... 164

(16)

3.54 sec-Butyl M ethyl Sulfoxide Tim e of Flight Mass Spectra (9R32 CO2

Laser Line, 0.1 W ) ... 167

3.55 sec-B utyl M ethyl Sulfoxide Mass S pectra (9R32 C O ; Laser Line, 0.1 W)168 3.56 sec-B utyl M ethyl Sulfoxide C O ; Laser O u tp u t Power D ependent Photol

ysis for Masses 56, 57 and 58 a m u ... 170

3.57 sec-Butyl M ethyl Sulfoxide C O ; Laser O u tp u t Power D ependent Photol

3.58 sec-Butyl M ethyl Sulfoxide C O ; Laser O u tp u t Power D ependent Photol

(17)

A ck n o w led g em en ts

This project was like a jo u rn ey th a t you walked a long distance to reach the final destination. B arriers were constantly encountered and decisions were m ade as to which direction to take to overcome the various problems. T hen when th e journey was finally over, you realized th a t you have ended up in Los Angeles when you really w anted to go to New York. W hen I first started this p roject, my goal was to use laser induced fluorescence to characterize the photofragm ent species but we never got around to this because of the numerous problems we encountered so th e product fragm ents were chareicterized with a laser ionized tim e of flight mass spectrom eter.

F irst of all, I would like to thank my supervisor. Professor Terry Gough for in troducing m e to th e fascinating field of spectroscopy when I was an undergraduate stu d en t a t th e University of W aterloo. Over th e past few years I have learnt a lot of inform ation and, for th a t, I shall be forever grateful.

N ext I would like to th an k Dr. Jack Barnes and Dr. W ayne Ingham for their assis tance w ith m y project. U nfortunately, the m ajority of th e synthesis work perform ed by W ayne was not used for this thesis. However, a valuable discussion w ith Wayne directed me tow ards investigating the skeletal rearrangem ent products after we had sta rte d the work w ith the sulfoxide molecules. Jack provided a lot of assistance with his skill and knowledge on designing and building electronic circuits, m achining and glcissware. Thanks to Jack, I did not have to go to th e M achine Shop, the Glass Shop or th e Electronics Shop very often. However, I still had to get some help from

(18)

Roy B ennett and Dick Robinson in the M achine Shop, Dave Searle and Sean Adams from the Glass Shop and Bob Dean and Terry W iley in the Electronics Shop and so I would like to th an k them as well. 1 would also like to thank G arth Irwin for the NM R sp ectra he recorded for me for dim ethyl sulfoxide and m ethyl phenyl sulfoxide. T h e 70 eV electron im pact mass spectra were recorded by Dave McGillivray.

I would also like to thank my fellow graduate students: Tangyu W ang, Terry Row at, Roy Jensen and Marcell Stoer. Numerous discussions with Tangyu and Terry provided additional guidance during the early stages of my project. Roy was of some help to me when he was an undergraduate stu d en t but even more so over th e past year since his retu rn from his voluntary te rm of exile in Denver. My friendship with M arcell goes back to our undergraduate days a t th e U niversity of W aterloo and we have been helping each other ever since. Felisa Stoer wrote th e first version of the com puter program m e to collect d a ta from th e LeCroy 9450 digital oscilloscope which Weis subsequently modified by Marcell. T he program m e th a t I w rote to interface with th e digital oscilloscope th a t replaced the analog switching box was too slow and so I

decided to use the program m e th a t they w rote and 1 would like to thank b o th of them .

Also, th e various discussions we have had over th e years helped me trem endously so th a t I could re-focus my energy on m y project.

Finally, I would like to th an k my parents, my sister Carol and m y b ro th er Frank for th eir encouragem ent over the years. W ith o u t their support during th e m any rough

(19)

(20)

C hapter 1

In trod u ction

C hem ists have always been interested in observing, explaining an d controlling chem ical reactions. O bserving a chemical reaction could be as sim ple as noting a colour change or detecting th e distinct odour of a particular product. U nfortunately, quali tativ e observations are not as useful zis quantitative results because m ore inform ation can be obtained by investigating num erical trends in th e d ata. For exam ple, by study ing th e ra te of form ation of a particu lar product species, one could determ ine if the reaction proceeded in a linear relationship (usually referred to as first order) or a q u ad ratic dependence (second order).

A g reater understanding is obtained by being able to explain th e outcom e of a reaction. J u st because a reaction is classified as being first order, it does not m ean th a t everything is known about th e reaction. A first order reaction in only one species which also has no dependence on any other species could be a unim olecular reaction.

(21)

Since a stable molecule will not spontaneously react to form products, this leads to the question, why did the molecule form the new product species? In order to explain the result of a reaction, scientists first propose a hypothesis to explain the reaction and if th e observations agree with the predicted results, it becomes a theory describing the experim ental process. Some of th e theories th a t were used to explain unim olecular reactions will be considered later in this chapter.

T h e n ex t logical step after being able to explain a reaction is controlling it so th a t a desired outcom e is achieved. A part of a molecule which behaves as a separate unit and in a certain way is cltissified as a functional group. Sometimes chemists want a reaction to occur with one functional group bu t not with another one on th e sam e molecule and so by preparing the molecule in an appropriate m anner, a site-specific reaction could then occur.

O ne m ethod of controlling a chemical reaction involves breaking a specific bond in a molecule. If a bond is considered to behave like a spring, one could (in principle a t least) p u t enough energy into th e vibrational m ode associate with the bond so th a t it will ru p tu re. Unfortunately, m ost of th e vibrational modes are not zissociated w ith ju s t one bond bu t describe a concerted m otion of th e molecule as a whole. In order to localize th e vibration to a single bond, a linear com bination of all of the modes needs to be considered such th a t m otion of the o th er atom s not involved in th e bond is cancelled out and th e only one which changes in length is the desired bond.

(22)

interactions of th e various atom s when th e molecule moves according to th e different vibrational modes. T he potential energy surface can be visualized as being like a m ountain range w ith peaks and valleys. Each valley represents a stable stru ctu re and the different valleys correspond to different isomers of th e molecule. A plateau can be thought of as a dissociative product channel. A molecule in one stable stru ctu re undergoing a chemical reaction could then follow a variety of different product chan nels, some of which lead to different isomers while others lead to th e dissociation of one or several bonds.

Although vibrationally exciting a molecule can result in th e system entering a product channel which corresponds to cleavage of a single bond, various unexpected product channels can also become energetically accessible and can also be entered by th e excited molecule. By studying the various species of th e different product channels when some condition is changed (usually th e am ount of energy absorbed by th e molecule), an understanding of the com peting reactions and th e ir respective reaction rates or likelihoods can be obtained.

From th e title of this thesis, it should be clear th a t this work involved a study of some sm all (i.e. less th a n twenty-five atom s) organic sulfoxide molecules w ith an infrared laser and it is understood th a t some form of detection system was employed. T he aim of this chapter is to provide background inform ation on these three com ponents of th e experim ent by first providing a general review of mass spectrom eters, then about laser ex citatio n and finally to discuss th e previous m ass spectrom etry and

(23)

photolysis work perform ed on these sm all sulfoxide molecules. C h ap ter 2 outlines the experim ental apparatus used in this work. T he experim ental results are presented in ch ap ter 3. T he Izist ch apter provides some concluding rem arks on this work and explains how it enhances th e picture presented by the previous work.

1.1 M a ss S p e c tr o m e te r s

Since the aim of this section is to provide a general review of mciss spectrom eters, an overview of two common instrum ental techniques p ertin en t to this work will be presented. A more detailed description of these and other techniques is beyond the

scope of this work but can be found in m any other places (reference [1] to nam e ju st a

single source). Because mass spectrom eters detect fragm entation species in addition to (or sometimes instead of) the parent molecule, unim olecular reaction theories can assist in providing an understanding of the mechanisms involved in form ing the various products. Consequently, this section will conclude with a discussion of this im p o rtan t topic.

1.1.1 Instrumentation

Mciss spectrom eters have three basic components: sam ple delivery system , m ass sep aratio n u n it and the detection system . T he sam ple can be delivered into th e mass sp ectrom eter in several different ways. For a gaseous sam ple, an effusive source could

(24)

be used to leak th e sam ple into th e mass spectrom eter a t the desired location. A pulsed m olecular beam w ith a carrier gas could be used to increzise th e m olecular density in th e mass spectrom eter. Non-gaseous sam ples could be inserted into th e m ass spectrom eter by placing th e sam ple on a probe and then positioning the probe a t th e desired location inside th e system .

Before covering the mass separation unit, some m ethods of detecting the sam ple will be covered. In some of th e first meiss spectrom eters, the detection system was a photographic p late [2-4]. A lthough this did provide an indication of which masses were present over a wide range, it was very difficult to q u an titativ ely d eterm ine how m uch of a certain species was observed. Replacem ent of the photographic plate with a position sensitive detector would provide a com plete spectrum where th e inten sity for a certain mass could be recorded. A more common su b stitu tio n is to use a single channel detector instead of the photographic plate and vary an instrum ental p a ra m ete r to generate th e desired mass spectrum .

Mciss separation of th e sam ple represents th e heart of a mass spectrom eter. By ionizing th e sam ple, separation can be achieved because of th e unique mciss-to-charge ratio for the ions of a certain mass. T he ionization process m ost com m only occurs by colliding the neutrally charged molecule w ith a highly energetic electron (usually aro u n d 70 eV of kinetic energy). T he neutral molecule absorbs some energy from th e electron an d em its an electron to become an ion. Colliding th e n eu tral sam ple m olecule w ith a positively charged ion is another way of producing a charged species

(25)

and this m ethod is referred to as chemical ionization. A nother common technique involves using a laser to photoionize th e sam ple. In all o f these cases, the ionization process occurs w ithin an electric field so th a t once the sam ple has been ionized, the ions are accelerated towards th e méiss analyzer.

M agnetic Sector D eflection Analyzers

T he m ost comm on m ethod of separating the different mzisses uses a m agnetic sector. T he ion beam is aligned to travel through the m agnetic sector such th a t the tra je c tories o f th e different masses are deflected by a different am ount. Thus, for a sector w ith a specific radius of curvature and a certain m agnetic field, only th e ions with the ap p ro priate kinetic energy will travel through and reach the detector. Varying any one of these th ree param eters (radius of curvature, m agnetic field or kinetic energy) while fixing th e o th er two will generate a mziss spectrum .

O ne problem w ith using only a single analyzer is th a t the ions of a certain mass can have a sm all range of kinetic energies due to the spread in the initial kinetic energy of th e neutrally charged sam ple. The spread in kinetic energies results in a range of ap p arent masses for th e ions which leads to a lower m ass resolution for the instrum ent. If th e ion beam travels through an energy analyzer before th e m agnetic sector, the ion b eam will be dispersed from th e central trajectory. A slit placed between the energy analyzer and th e m agnetic sector would then narrow th e range of kinetic energies of th e ions entering th e m agnetic sector by only allowing a narrow range of trajectories

(26)

to pass from th e energy analyzer and into th e m agnetic sector. T his m ethod improves the m ass resolution b u t at th e expense of decreasing the num ber of ions which reach th e detector.

Stable ions will travel from th e ionization region through th e analyzers and on to th e detector w ithout fragm enting. However, some ions will dissociate within the analyzers. This results in a broad peak in the mass spectrum a t a different mass. Even though these m etastable ions com plicate the mass spectrum , they do provide some valuable inform ation. T he ion will have an apparent mass of m* given by [5]

= (1.1)

TTZp

where nip represents th e m ass of the unstable parent ion form ed in the ionization

region and is th e mass o f th e daughter ion formed while the p arent ion was within

the analyzers. T he kinetic energy of the ion is given by

eV = ^rripv^ (1.2)

while th e m agnetic deflection of the ion is for the daughter ion and is a balance betw een th e cen trip etal force and th e centrifugal force. T he velocity for the ion is then

(1.3)

(27)

where e is th e charge on th e ion and H is th e m agnetic field strength. S u b stitu tio n of this te rm for the velocity into th e kinetic energy equation w ith some additional rearrangem ents results in

I'"

or

The presence of a m etastable peak in th e mass spectrum thus provides a m echanism describing th e form ation of the daughter ion from the parent m olecular ion.

T im e o f Flight Analyzer

A tim e of flight analyzer involves accelerating the ions from th e ionization region towards th e d etecto r and achieves mass separation because all of th e ions will have the sam e kinetic energy so ions of dissim ilar masses will have different velocities. Hence, th e tim e for ions of dissim ilar masses to travel through a drift region of a fixed length will be different. Wiley and M cLaren improved th e resolution over a single accelerating region tim e of flight m ass spectrom eter by adding a second accelerating

region right after th e ionization region [6]. T h e tim e of flight m ass spectrom eter used

(28)

and M cLaren an d th e specifics axe covered in m ore detail in section 2.3 startin g on page 53. O ne advantage w ith using a tim e of flight m ass analyzer instrum ent is th a t a com plete m ass spectrum can be obtained in a few microseconds. This would be an advantage in a system where the sam ple m olecular density changes unexpectedly. A m ajor disadvantage w ith a tim e of flight mass spectrom eter is th e lower mass resolution th a n with a m agnetic sector analyzer.

The spread in the initial kinetic energy of th e neutrally charged sam ple molecules is also a problem w ith a tim e of flight mass analyzer. The extrem e situation involves considering tw o ions w ith the same initial kinetic energy, one initially moving towards th e detector a n d th e o th er moving away from the detector. T he ion moving away from th e d etecto r will be decelerated until it stops and th en th e ion will be accelerated tow ards th e d etecto r. W hen the ion returns to th e original position, it has the same kinetic energy as it initially had but it is now travelling tow ards the detector. The tim e th a t it takes th e ion to retu rn to the original position is referred to as the tu m -aro u n d tim e and th e total flight tim e for th e two ions will differ by this am ount.

T he problem of th e turn-around tim e can be corrected in several different ways. O ne m ethod of dealing w ith this problem involves increasing th e stren g th of the field in th e ionization region. T h e tim e spread is decreased because th e stronger field causes th e ion travelling away from the detector to rapidly tu rn around and be accelerated in th e opposite direction. A nother m ethod involves lengthening th e drift region. This increases th e to ta l flight tim e and so th e relative am ount of tim e for th e ion to tu rn

(29)

around (com pared to th e to ta l Sight tim e) is decreased.

T h e mass resolution can also be improved by adding an elem ent to deal w ith the spread in th e initial kinetic energies and one common type of in strum en t which does this is referred to as a reflectron tim e of flight mass spectrom eter [7]. An ion m irror uses an electric field to decelerate th e ion and then accelerate it again, usually in a different direction. Ions w ith a slightly higher kinetic energy will travel further in the ion m irror before being deflected away from the m irror. This results in a compression of the tim e spread due to th e range of initial kinetic energies for ions of th e same mass because the faster ions (w ith the larger initial kinetic energy) travel a longer distance from the ionization region to the detector.

1.1.2 Unimolecular Reaction Theory

As previously m entioned, mass spectrom eters require ionizing th e sam ple b u t, un fortunately, th e ionization process can result in a parent ion th a t is unstable which can th en dissociate. This type of behaviour is designated as a unim olecular reaction because it involves a single molecule reacting to form products. T he form ation of the fragm entation species in a mass spectrum could then be tre a ted as being governed by rate processes of unim olecular reactions. Although this type of calculation was not perform ed for the work presented in this thesis, a historical overview of unim olecular reaction theories is provided because of the im portance of this topic to th e field of m ass spectrom etry.

(30)

T he Radiation H ypothesis

T he first som ewhat successful a tte m p t a t explaining th e m echanism behind unimolec ular reactions was th e radiation hypothesis developed prim arily by Perrin in 1913

and further refined after W orld W ar I [8]. It stated th a t the energy necessary for

th e molecule to undergo either a rearrangem ent (i.e. isom erization) or a decompo sition reaction was provided by th e environm ent th a t the molecule was located in and the energy was in th e form of a single photon of radiation. Some people have interpreted this to m ean th a t the radiation m ust come from the walls of the container. T he justification for this hypothesis wzis th a t unim olecular reactions followed a linear dependence on the pressure of the system , not quadratic as would be expected for bim olecular reactions which involve collisions to energize th e molecule.

U nfortunately, at th e tim e th a t Perrin and other researchers were spending so much effort refining the radiation hypothesis to fit the available d a ta , there was no known reaction th a t was truly unimolecular; the reactions th a t were believed to follow a unim olecular mechanism were later shown to be complex in nature. One such exam ple was th e decom position of N jO s according to th e reaction

2 W 2 O 5 —> 4 N O 2 + O 2 ( 1 . 6 )

investigated by Dziniels and Johnston [9,10]. T he experim entally observed ra te of

(31)

so th e au th ors believed th e y had found a unim olecular reaction. Subsequent work determ ined th a t the decom position was explained by the following m echanism [11-14]

N2O5 ^ N O2 + N O3 (1.7)

N O2 + N O3 N2OS (1.8)

k2.

NO2 + NO3 —y NO2 + O2 4" N O (1.9)

N O + N2O5 ^ 3NÜ2 (1.10)

w ith the ra te of N2O5 consum ption given by [15]

d [NgOg] 2k y k2

dt fc_i + At2

[N2O,]. (1.11)

T h e rate of N2O5 consum ption denoted by equation 1.11 is first order but th e decom

position o f N2O5 does not occur by an elem entary unim olecular reaction.

One of th e opponents of the radiation hypothesis was Langm uir who pointed

ou t th a t th ere was no direct evidence supporting this th eo ry [16]. He believed th a t

unim olecular reactions had to be explained in term s of q u an tu m mechanics. Langmuir s ta te d th a t every chemical reaction was a discontinuous process, going from reactants to products. His m ain objection to th e rad iatio n hypothesis was th a t th e refinements th a t were necessary in order to explain th e experim ental evidence com plicated the sim ple p ictu re of unim olecular reactions to th e point o f m aking this theory ra th e r absurd.

(32)

T h e chief d e tra cto r of th e rad iatio n hypothesis was Lindem ann who pointed out th a t for th e inversion of sucrose, th e radiation hypothesis did not predict th e correct relative rate of reaction in sunlight com pared with in darkness [17]. T he intensity

of 1.05 f i m light is 5 x 10^^ tim es greater in sunlight th a n in th e dark and so should

correspond with a m uch higher rate in sunlight while th e experim ental evidence sug gested th a t the reaction proceeded a t about the sam e rate w ith or w ithout sunlight.

Lindem ann later pointed o u t an inconsistency between the supporters of th e radia tion hypothesis [18] when Perrin suggested th a t the ra te of inversion of sucrose was

dependent upon th e absorption of sunlight in several bands, not ju st a t 1.05 fxm [19]

while Lewis had found a single band which supported the position o f th e radiation hypothesis in this reaction [20].

T he Lindemann M echanism

Instead of ju s t arguing against th e radiation hypothesis, Lindem ann also proposed th a t th e molecule was energized by a collision with another molecule before a uni m olecular reaction occurred [17]. This could be accounted for w ith the following m echanism

A + A % A ' + A (1.12)

A "-p A —> A - { - A (1.13)

(33)

where A “ represents a molecule with sufficient energy to surpeiss th e barrier to reac

tion. The rate of producing the energized species is while th e ra te of consuming

th e energized species is th e sum of A:_i[v4*][A] and

For high pressures, th e rate of de-energizing A ’ (A;_i[A*][A]) will be much g reater

th a n the rate of pro d u ct form ation (fcalA*)). In this situation, th e ra te of forming the

energized species is equal to th e rate of de-energizing Thus,

k^[A]^ = (1.15)

or, in term s of solving for [A*],

[A*] = ^ [ A ] . (1 16)

The rate of product form ation is then equal to (A:ifc2/fc_i)[A] and is first order in [A].

Thus, a collisional energization m echanism could account for th e observed first order behaviour of unim olecular reactions.

For low pressures, the rate of de-energizing A’ is much sm aller th a n the ra te of product form ation. T he rate of product form ation is then equal to th e ra te of form ing th e energized species and is second order (fci[A]*). This result was a significant change from the radiation hypothesis which stated th a t unim olecular reactions were first order at all pressures. The second order dependence a t low pressures was exper im entally verified in 1927 by R am sperger [21] and effectively closed th e door on th e

(34)

rad iatio n hypothesis.

An a lte rn a te derivation of the unim olecular reaction ra te involves using th e steady- sta te approxim ation, which states th a t the change in the concentration of interm ediate species over tim e is equal to zero (i.e. the concentration, w hatever it m ight be, does not change significantly). This approach yields

and th e ra te of product form ation is

A t high pressures, is much larger than 6 2 and th e reaction ra te becomes

k i h

(1.19)

k .

ju s t as before. At low pressures, At2 is much larger th a n A:_i[A] and th e reaction ra te

simplifies to

k ,[A]^ (1.20)

which, again, is th e sam e result as was previously obtained.

(35)

order ra te coefficient ( k ) by fc[A], th e following expression is obtained by equating th e a p p aren t ra te of th e reaiction to th e rate of product form ation

or

k — f^ik2/ k - i /I 2 2)

-- 1 -h k , /( / :_ .[/I])' ( ^ ^

An alternativ e representation, which is more useful, is to consider th e reciprocal of k

A plot of 1/[A] against i / k should yield a straight line w ith slope 1/fci and intercept

k - i / { k i k2). U nfortunately, experim ental results show a significant deviation from this

expected linearity and this problem m ust be addressed by a successful theory.

Approach M ade B y H inshelwood

In 1927, Hinshelwood refined Lindem ann’s m echanism by allowing th e energy avail

able to th e molecule to be d istrib u ted among th e s vibrational modes of th e m ole

cule [22]. Since each vibrational mode involves two q u ad ratic energy term s, the

(36)

obtained

where k s is the B oltzm ann constant and T is the tem p eratu re in kelvin. T he rate

coefficient for a reaction is taken to be proportional to th e fraction of molecules above

th e threshold energy so is given by

h . - L

k ^ i 1

where it is assum ed th a t there is no energy barrier for th e de-energizing reaction and so th e fraction of molecules above th e threshold energy is unity.

A lthough this expression provides much higher rates for th e activation reaction and b e tte r agreem ent with experim ental results, Hinshelwood’s approach still has some problem s. T he m ost significant problem wais th a t only about half o f th e num ber of vibrational modes needed to be considered to provide agreem ent w ith experim en ta l data. A nother problem was th a t / , the fraction of molecules above th e threshold energy, introduced a term w ith a strong tem p eratu re dependence b u t th ere is no ex perim ental evidence to support this claim . Also, ju s t as w ith L indem ann’s treatm en t,

(37)

R ic e - R a m s p e r g e r - K a s s e l T r e a tm e n t

A problem w ith the treatm en t m ade by Lindem ann and Hinshelwood is th a t the apparent first order ra te coefficient given by

*

= 1

involves ra te coefficients which are independent of the am ount of energy in the ex cited molecule. The energizing reaction rate coefficient (fci) should increase as the energy content is increased because the m ore energy a molecule has, the greater the

probability th a t a collision will produce an energized molecule. Thus, k i / k - i should

be dependent upon the energy content of the molecules.

T he ra te coefficient for th e product forming reaction (fej) should also be depen dent upon th e energy content. The probability th a t some internal m ode will contain enough energy to surpass the reaction barrier will increase as the energy content of th e molecule is increased. An im provement to Lindem ann and Hinshelwood’s theories

would, therefore, tre a t k i / k - i and as being dependent on th e energy.

In 1927, Rice and Ram sperger published a p aper which allowed the energy to be distrib u ted am ong th e various individual energy term s of th e molecule [23,24]. A re action would occur when th e critical term surpassed some threshold energy. Shortly thereafter, K assel published a sim ilar treatm en t b u t he allowed th e energy to be dis trib u ted am ong the various vibrational modes of th e molecule [25]. This classical me

(38)

chanics treatm en t of the energy distribution was followed up w ith a sim ilar quantum mechanical treatm en t [26]. Because the tre a tm e n t m ade by Rice and Ram sperger was so sim ilar to th a t m ade by Kassel, people refer to this as the R RK theory of unim olecular reactions.

T he expression for k i / k ^ i is obtained by startin g with equation 1.25 and allowing

th e energy in the critical mode to vary above th e threshold value. This yields

where E " refers to an am ount of energy above the m inim um value. T he energy-

dependent product form ation rate coefficient is taken as being proportional to the probability of the reaction occurring. This is the ratio of th e num ber of ways of distributing the energy above th e threshold value in all of the oscillators to the num ber of ways of d istributing all of the energy in all of the oscillators. Using this approach for both th e classical and qu an tu m mechanics treatm en ts results in a reaction probability given by [25,26]

■ E - E o V - ^ -

( ^ )

(39)

R ic e - R a m s p e r g e r - K a s s e l- M a r c u s ( R R K M )

T h e Rice-Ramsperger-Kcissel tre a tm e n t was extended by Marcus and Rice in 1950 [27, 28] by including tran sitio n -state theory and once again in th e early 1960s by W ieder and M arcus [29] where a q u an tu m mechanical approach wzis incorporated. T he reac tion schem e is given by

A + M A '- j- M (1.29)

A *-f M A -t- Af (1.30)

A" A* (1.31)

A* { P r o d u c t s ) (1.32)

w here th e ra te coefficients for the energization step { k i) and th e activation step

{ k a ( E ' ) given by reaction 1.31) are b oth taken as being energy-dependent. It can be shown th a t [30-32]

where h is P lanck’s constant, N * { E ' ) is the density of states a t energy E ' , E}^ is

th e energy in the vibrational and rotational modes of P ( E ^ ) is the num ber of

vibrational-rotational states of A* w ith energy exactly equal to and th e sum m a

(40)

ra te coefficient can th en be shown to be

t 1 /■■” ( E P i . E * ) ) e x p ( - E + / t , T ) d E *

^ " ~ h q i J E , l + K ( E - ) l k . , [ M ] '' '

Since th e extension of th e Rice-Ramsperger-Kassel model m ade by M arcus is so de pen d en t upon RRK theory, it is commonly referred to as RRK M theory to honour all four contributors. The success of RRKM theory in estim atin g unim olecular reac tion rates over the past fifty years has led it to become th e stan d ard theory used to describe unim olecular reactions.

1.2 L aser E x c ita tio n

In th e previous section, an overview of some of the theories describing unim olecular reactions was presented in order to provide an understanding of w hat happens when an energetically excited molecule fragm ents. T he work presented in this thesis used

a CO2 laser to vibrationally excite a molecule still a t th e electronic ground state as

opposed to m ost of the other photolysis work reviewed in th e next section which uti lized a vacuum ultraviolet Iziser to raise th e molecule to an excited electronic state. Staying a t the electronic ground sta te simplifies th e analysis of unim olecular reac tions because a change in th e electronic level does not need to be considered. This section sta rts w ith a general discussion on how a m ultiple num ber of objects (atom s, molecules or photons) in teract to form some other molecule (in th e czise of atom s and

(41)

molecules) or an excited species (when an ato m or molecule absorbs several photons). This discussion is of some benefit when laser initiated unim olecular dissociation is considered. In addition, an overview of infrared m ultiple-pboton dissociation is pro vided to connect the two other subsections together with th e apparatus used for this work.

1.2.1 Interaction of Multiple Objects

This section deals w ith the interaction of m ultiple objects through some type of reaction. T he objects could all be molecules (or molecules and atom s) in which case the interaction is a collision. T he other possibility th a t will be considered is the case where one atom or molecule interacts w ith an electric field from a laser beam . In this case, th e o th er objects are photons and the interaction is an absorption process. The subsequent paragraphs are left a bit general to deal with these two possibilities. T he section will end with a treatm en t of the differences in the analysis due to the separate cases.

If one considers th e interaction of one species, generically designated as A, with a

m ultiple num ber of another species, B , the process where species A absorbs a m ultiple

num ber of species B m ay be represented by

(42)

If this chemical equation represents an elem entary process, th e rate law for this reac tion would be

V = k[A][B]'^ (1.36)

and the reaction would be of order n in species B and overall order (n -|- 1).

T h e interaction represented by equation 1.35 involves (n -|-1) species b u t th e like lihood of (n -f 1) species sim ultaneously interacting, under norm al circum stances, decreases rapidly as n increases. A more likely scenario involves a sequential interac tion as denoted by th e following reaction scheme

A + B A B (1.37)

A B + B % A B -i (1.38)

A B ^ j . i ) + B % A B , (1.39)

A B ( n - i ) + B ' % (1.40)

and th e ra te law for th e equation where j copies of B have been absorbed by the

com plex is

(43)

Each of these chemical equations represents an elem entary bim olecular process which

is first order in species B and second order overall.

T he ra te law for equation 1.40 is

t; = (1.42)

and it involves the in term ed iate species AB^^n-iy If th e stead y -state approxim ation

is used, th en the rate o f form ation of this species equals the ra te of depletion which is depicted m ath em atically as

= 0 (1.43)

or

(1 4 4 )

(44)

C ontinuing this process of dealing w ith the various interm ediate species yields

v = ki[A][B] (1.46)

which is first order in species B and second o rder overall. This result is vastly different

from the case where the reaction was treated as a single elem entary reaction which

was of order n w ith respect to B and order (n + 1 ) overall.

The problem with using the steady-state approxim ation is th a t it is possible th a t not all of an interm ediate is used to m ake the next product. In the case where more of an interm ediate is produced th an consumed by a subsequent reaction, the rate of form ation would be

> 0 (1.47)

dt

which yields

T he inequality can be re-arranged to give

(45)

and su b stitu tio n into equation 1.42 results in

ü = (1.50)

R epeating this su b stitu tio n procedure for th e other interm ediates yields

V < ki[A][B]. (1.51)

T he apparent ra te constant m ust be less than k i and once again the order of the

reaction is clearly not n w ith respect to species B .

W hen species B is either an atom or a molecule, it is very rare th a t two (or more)

of them will collide sim ultaneously w ith species A because it is highly unlikely th a t

three (or more) species travelling in random directions will occupy the sam e volume of space at th e sam e tim e. In this case, th e sequential addition process predom inantly occurs. However, when species B is a photon, a strong electric field (as occurs w ithin a tightly focused laaer beam ) makes it possible for num erous photons to in teract w ith the absorbing species w ithin th e short period of tim e th a t the photons take to travel through th e interaction volume. In this fashion, th e net result is th a t it appears th a t a m ultiple num ber of photons have been absorbed by species A in a single step. This is referred to as a m ultiphoton absorption process. In addition to th e m ultiphoton process, th e sequential photon absorption process can occur when the interm ediate is a stable excited species w ith a finite lifetime. To em phasize th e difference in the

(46)

photon absorption m echanism , th e second case is occasionally referred to as m ultiple- photon absorption.

1.2.2 Infrared M ultiple-Photon Dissociation

The previous section dealt with a molecule absorbing a m ultiple num ber of photons. O n its own, this process has a lim ited appeal b u t it is very helpful in preparing the system for a subsequent reaction. In this section, this process is extended to th e case where enough photons have been absorbed so th a t the molecule dissociates. Because

th e work for this thesis used a CO2 laser to excite a molecule above the dissociation

threshold, a review of vibrational modes and infrared m ultiple-photon dissociation will be briefly discussed.

By m aking the approxim ation th a t a bond behaves like a spring, the vibrating molecule undergoes a restoring force which is proportional to its displacem ent. This vibration can th en be described as a harm onic oscillator. As such, it will have equally

spaced energy levels which are separated by h u where h is Planck’s constant and u

is th e frequency of th e vibration. A laser of th e sam e frequency could th e n excite th e molecule above th e vibrational ground state. V ibrational frequencies are usu ally around 500 to 3000 cm “ ‘ which is in th e infrared region of the electrom agnetic spectrum .

U nfortunately, th e sim ple picture of treatin g th e vibrational m ode as a harm onic oscillator is not correct. For a typical vibrational m ode, th e restoring force gets

(47)

weaker as the displacem ent increases. This results in the energy levels a t the larger displacem ents being closer together th a n with a harm onic oscillator. Consequently, a laser which can excite a molecule in th e ground s ta te by the absorption of a single photon m ay not be able to further excite the molecule.

O ne way in which further excitation can be accomplished is by the absorption of a m ultiple num ber of photons as m entioned in the previous section. U nfortunately, the probability of absorbing a m ultiple num ber of photons in a single step is proportional to th e laser intensity raised to th e power of the num ber of photons (i.e. absorbing three photons would be proportional to the cubed power of the laser intensity) [33]. Since th e dissociation energy of a polyatom ic molecule is typically equivalent to ab o u t 20-30 infrared photons [34], this would require a laser intensity on the order of a gigawatt per square centim etre [35].

A lthough this intensity is attain ab le, ef&cient dissociation is possible at m uch lower intensities (about 5 M W /cm ^ [36]) which suggests th a t a different m echanism m ust be responsible for th e dissociation process. T he anharm onicity resulting from the weakened restoring force a t larger displacem ents m ay also remove the independence of th e norm al vibrational modes. This coupling of vibrational modes allows the molecule to redistrib u te th e energy from one m ode into some of th e other modes. For exam ple, consider the case of three vibrational modes which are coupled together and where the frequency of one mode is equzd to th e sum of two other modes (i.e.

(48)

energy from th e first m ode into one quantum in both of the o th er two modes.

It should be pointed out th a t th e coupling of vibrational modes is not as simple as was described in the previous paragraph. A t low levels of excitation, anharm onicity plays a m inor role and so th e energy content tends to rem ain localized in the excited vibrational mode. O ne common m ethod by which molecules redistribute energy to different vibrational modes is by a collision with another molecule. At higher levels of excitation, anharm onicity term s become im portant and an isolated molecule can red istrib u te energy am ongst the various vibrational modes w ithout collisions [37]. This intram olecular vibrational energy redistribution is quite fast (1-10 ps [38]) and usually occurs before th e molecule dissociates.

This now provides a picture of the current model for the m ultiple-photon absorp tion process. Initially th e excitation occurs in a single vibrational mode, but due to th e anharm onicity, th e molecule can only be excited a few levels above the ground state. Because of the coupling o f the vibrational modes, energy can be redistributed from the absorbing m ode and transferred into other modes. This allows the absorb ing m ode to be excited once again and the absorption-redistribution cycle can be repeated m any tim es while the molecule is still in the electric field of th e laser beam.

As m ore vibrational energy is put into th e molecule, the num ber of perm utations in how th e energy can be distribu ted into th e various vibrational modes increases. Since each p erm u tatio n corresponds to a particulzir sta te of th e molecule, the density of states a t a particu lar energy level increases as m ore vibrational energy is put

(49)

into the molecule. So far, only th e case where the states are a t exactly the sam e energy level has been considered. Because of the frequency m ism atch o f the different vibrational modes, there will be some states a t a slightly different energy level. For low levels of vibrational excitation, this energy gap will be quite large b u t the energy gap will becom e narrower as th e molecule absorbs more vibrational energy. This results in an alm ost continuum of energy levels for the molecule. Once the molecule has been excited into this quasi-continuum , it can easily be further excited up to the dissociation threshold by th e absorption of several photons, a single photon at a tim e. A nother way o f looking at this is th a t an increasing density of states m eans th a t the probability of th e current energy level in the quasi-continuum being separated from

a higher level by hi/ increases th e further the molecule has been excited up into the

quasi-continuum . This, in tu rn , m eans th a t th e likelihood of a photon being absorbed by the molecule increases. Consequently, the molecule can easily be excited up to the dissociation threshold by th e absorption of a few photons a t a tim e once it is in the quasi-continuum of energy levels.

T he molecule can still be excited above th e dissociation threshold but now two com peting processes have to be considered: th e molecule being fu rth er excited above th e lowest dissociation energy threshold and th e molecule fragm enting. The dissocia tion process can be thought of «is ru p tu rin g a single bond in the molecule. In order for th e bond to break, th e corresponding vibrational mode(s) m ust be excited above the dissociation threshold. B ut because th e modes are coupled together, th e to ta l

(50)

vibra-tionaJ energy of th e molecule when it dissociates is usually larger th an th e am ount needed to reach this threshold. In other words, usually other vibrational modes are also excited above their ground state when th e dissociating m ode h«is enough energy for th e molecule to fall ap art.

1.2.3 Laser Initiated Unimolecular Dissociation

As previously m entioned, one of the first theories to try and explain unim olecular reactions was th e radiation hypothesis (see section 1.1.2 on page 11). It essentially s ta te d th a t th e molecules obtained enough energy to react by absorbing radiation from th e surrounding environm ent. A t the tim e, this was only possible by therm ally exciting th e molecules. Because the am ount of th erm al energy available was generally insufficient to cause a reaction, the radiation hypothesis was dism issed. A nother com p etin g theory was th a t th e molecules obtained th e required energy through collisions w ith o th er molecules. Strictly speaking, collisions involve a t least two molecules and so, technically, are not unimolecular b u t bim olecular reactions. However, as was pre viously shown, at higher pressures these reactions involve only th e excited molecule an d exhibit a unim olecular behaviour. It is interesting to note th a t th e im p o rtan t point of b oth th e radiation hypothesis and th e colllsional explanation of unim olecular reactions is th a t th ey both require th a t the reacting molecules o b tain enough energy to surpass an energy barrier. W ith th e developm ent of high power lasers, it is now possible to stu d y tru e unim olecular reactions by exciting th e molecules w ith rzidiation

(51)

from th e laser.

U ndergraduate Physical C hem istry textbooks sta te th a t th e ra te of a reaction involving th e absorption of a photon is given by th e num ber of q u a n ta absorbed per

u n it tim e p er unit volum e {labs) [39,40]. W ith this in m ind, consider th e following

reaction scheme: A -|- hi/a Igb^ A* (1.52) A + M % A *-f M (1.53) A - + M % A + M (1.54) A ' % A + hua (1.55) A* % A* (1.56) A* % B + C (1.57)

where M is some a rb itrary bath gas molecule and B and C are the fragm entation

products.

T h e second and th ird reactions account for collisional activation and de-activation

of A . T he fourth reaction represents the excited A molecule em ittin g a photon to

de-energize A ' . Ideally, these three reactions should be negligible b u t are included

(52)

T he ra te of change of [A*] is given by

^ = Iabs[A] + k2[A][M] - k , [ A ' \ m - k M ' ] - (1.58)

b u t because [A*] is a reaction interm ediate, it is used up as soon as it is form ed and so by the steady-state approxim ation the rate of change of [A*] should equal zero. This results in

14-1 - + f<=2[A][M] .

k3[M] + k, + k, •

^ ’

Similarly, [A*] is given by

[A>] = ^ [ A - ] (1.60)

6

which yields

f i t ] — ^5 fo6»[A]

+

[A] [ Af ] . gi \

^ " Are

k,[M] + k^ + ks

' ^ ‘ ^

T he rate of change of product B is

d [B ]

= k e W (1.62)

(1.63)

dt

^ sfa fc ifA ] - f

k2ks[A][M\

(53)

At low pressures, [M] —> 0 and

(1.64)

dt + ks

and is first order in [A] as required.

A lthough the final result presented in th e previous paragraph was dependent upon th e absorption of a single photon (equation 1.52 on page 32), th e sam e result is obtained if this equation is replaced by

A + n hi/a —>■ A " (1.65)

where th e excitation is considered to occur by a m ultiple-photon absorption process. In section 1.2.1 (startin g on page 22), it was shown th a t the ra te law for the sequential absorption of n photons was

V = ki[B][A] (1.66)

which becomes

V = kiIabs[A] (1.67)

w ith th e n o tatio n used in this section. Since this is the sam e expression as was used for th e developm ent of equation 1.63, it stands to reason th a t th e sam e result was

(54)

obtained when eith er equation 1.52 or 1.65 was used in th e laser in itiated unim olecular dissociation reaction scheme.

1.3 S m a ll O rgan ic S u lfo x id e M o le c u le s

As m entioned in th e previous section, this work utilized a C O3 laser to vibrationally

excite a molecule in its ground electronic state. T he m ajo r laser lines are around

9.6 and 10.6 f i m (1040 and 940 cm “ ‘) and so whichever molecules are selected, they

m ust absorb near one of these two bands. T he original objective of this project was to vibrationally excite transition m etal carbonyls to stu d y th e fragm entation m echanism s as the CO groups were lost from th e parent molecule. A lthough the single m etal atom carbonyl molecules were readily m ade, the molecules with a greater num ber of m etal atom s were m uch m ore diflBcult to synthesize. A t about this tim e, th e tim e dependent infrared laser photolysis stu d y on a few sm all organic sulfoxide

molecules perform ed by Gross et al. [41] was published an d it was decided th a t a

com plem entary investigation could be conducted by ou r group.

In th e 1960s, R ayner et al. proposed th a t therm al racem ization of a group of di-

aryl, alkyl aryl and dialkyl sulfoxides occurred by pyram idal inversion from a molec ular vibration w ithout breaking a carbon-sulfur bond [42,43]. V ibrational excitation of this cleiss of molecules should yield some interesting results due to the m otion of th e sidegroups relative to the sulfur a n d /o r oxygen atom s, m aybe even th e m igration of p a rt of th e molecule in some instances. Because th e experim ental work for this th e