Citation
Computational Investigation of 1-Fluorocyclopropenes

Material Information

Title:
Computational Investigation of 1-Fluorocyclopropenes
Creator:
SHELTON, G. ROBERT ( Author, Primary )
Copyright Date:
2008

Subjects

Subjects / Keywords:
Atoms ( jstor )
Carbenes ( jstor )
Carbon ( jstor )
Cations ( jstor )
Chemicals ( jstor )
Electronic structure ( jstor )
Fluorine ( jstor )
Geometry ( jstor )
Lead ( jstor )
Orbitals ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright G. Robert Shelton. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
12/31/2005
Resource Identifier:
436098607 ( OCLC )

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

COMPUTATIONAL INVESTIGATION OF 1-FLUOROCYCLOPROPENES By G. ROBERT SHELTON A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2004

PAGE 2

I dedicate this work to my family. Their st rength, encouragement, and love will never be forgotten.

PAGE 3

ACKNOWLEDGMENTS First and foremost I would like to thank my supervisory committee chair. If not for the patience, guidance, and flexibility of Dr. William R. Dolbier the time spent at University of Florida would have been wasted. With as much owed respect and thanks, I acknowledge Dr. Merle Battiste for his insight and inspiration on this project. Without Dr. Adrian Roitberg’s help and conversations, frustrations would have seriously shortened the breath of knowledge gained here at Florida. The friendship and mentoring of Dr. Khalil Abboud will not be forgotten. Personally I would like to acknowledge my family: Lindsey Averill, mother Susan, Grandparents “PaPa & DeeDee”, and my brother Sam. Without their support, love, and teachings I would not be in this position today. I thank them. Over the years I have had the privilege to be taught by some truly inspirational teachers. I extend my thanks to Ross Brewer for his teachings and for instilling in me the need to understand the people and events of history. The teachings of Keith Mitchell on how to find the beauty in music, nature, and people will never be forgotten. Finally I thank Dr. Tim Patrick for showing me the doorway into the realm of organic chemistry. I would like to take a moment to acknowledge friends, chemists, and fellow group members who have truly made the time here at Florida enjoyable. Thanks go to Randal Harms and Scott Przybysz for being great friends throughout the entire educational experience. Thanks go to Amy Austin, David Baker, Tom Cameron, and Tim Foley for introducing new types and ways of doing chemistry. Finally, I thank the present and past iii

PAGE 4

group members Marshall Baker, Tyler Schertz, Joe Cradlebaugh, Zhai, Merry, and Samia for making the time in the lab a true learning experience. iv

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................iii LIST OF TABLES ............................................................................................................vii LIST OF FIGURES ...........................................................................................................ix ABSTRACT ......................................................................................................................xii CHAPTER 1 FLUORINE SUBSTITUENT EFFECTS.....................................................................1 Introduction...................................................................................................................1 Structure, Bonding and Reactivity................................................................................3 Saturated Systems..................................................................................................3 Unsaturated Systems.............................................................................................4 Cyclopropane.........................................................................................................5 Structure.........................................................................................................5 Strain..............................................................................................................6 Electronic structure........................................................................................6 Fluorine substituent effect..............................................................................9 Cycopropene........................................................................................................11 Structure.......................................................................................................11 Strain............................................................................................................12 Electronic structure......................................................................................12 Fluorine substituent effect............................................................................13 Reactive Intermediates................................................................................................14 Carbocations........................................................................................................14 Free Radicals.......................................................................................................15 Carbene................................................................................................................17 2 THEORETICAL STUDY OF THE THERMOLYSIS OF 1-FLUOROCYCLOPROPENE.....................................................................................19 Part 1: The C 3 H 4 & C 3 H 4 F Initial Ring Opening........................................................19 Introduction.........................................................................................................19 The C 3 H 4 Surface: Ab Initio Study.....................................................................21 v

PAGE 6

Structures and stabilities of isomers.............................................................21 Thermal isomerization..................................................................................23 Computational detail...........................................................................................25 The C 3 H 4 Surface: DFT study.............................................................................27 Structures and stabilities of isomers.............................................................27 The C 3 H 3 F Surface: DFT Study..........................................................................30 Fluorines substituent effect..........................................................................30 1-Fluorocyclopropene..................................................................................32 3-Fluorocyclopropene..................................................................................36 Conclusions..................................................................................................37 Part 2: Difluorocyclopropenes; Computational Study of Structures and Initial Thermal Rearrangement.........................................................................................38 Introduction.........................................................................................................38 3,3-Difluorocyclopropene...................................................................................40 1,2-Difluorocyclopropene...................................................................................42 Conclusions.........................................................................................................43 Part 3: DFT Study of 1-Fluoromethylcyclopropenes.................................................44 Introduction.........................................................................................................44 1-Fluoro-2-methylcyclopropene..........................................................................47 1-Fluoro-3,3-dimethylcyclopropene....................................................................49 Conclusions.........................................................................................................51 3 2-FLUOROCYCLOPROPENYLCARBINYL RADICAL AND CATION REARRANGEMENTS..............................................................................................53 Introduction.................................................................................................................53 Cation..................................................................................................................54 Radical.................................................................................................................55 Computational Detail..................................................................................................56 The Cyclopropenylcarbinyl Cation: A DFT Study.....................................................56 2-Fluorocyclopropenylcarbinyl Cation: A DFT Study...............................................59 Gas Phase.............................................................................................................59 Solvent Phase.......................................................................................................63 Cyclopropenylcarbinyl Radical: A DFT Study..........................................................64 2-Fluorocyclopropenylcarbinyl Radical: A DFT Study.............................................66 Distal Ring Opening............................................................................................66 Proximal Ring Opening.......................................................................................68 Conclusions.................................................................................................................73 REFERENCES..................................................................................................................74 BIOGRAPHICAL SKETCH.............................................................................................78 vi

PAGE 7

LIST OF TABLES Table page 1-1 Electronegativity of several common atoms..............................................................2 1-2 Van der Waals Radii..................................................................................................2 1-3 Carbon-halogen bond lengths () over the BDE (kcal•mol -1 )...................................3 1-4 Experimental fluoromethane angles...........................................................................4 1-5 Bond distances, angles, and BDE of fluoroethenes................................................4 1-6 Heats of Hydrogenation.............................................................................................4 1-7 Structure parameters of cyclopropane........................................................................5 1-8 Bond lengths in different cycloalkanes......................................................................6 1-9 Geometric data for cyclopropene.............................................................................11 1-10 Wiberg’s isodesmic equations and computed results...............................................13 1-11 CC bond lengths(r/pm) for several cyclopropenes...................................................14 1-12 Radical stabilization energies (RSE) as described by the isodesmic equation.........17 2-1 Computed and experimental geometries of the most stable C 3 H 4 isomers..............23 2-2 Geometries of intermediates on the singlet surface.................................................23 2-3 DFT geometric data and comparison to previous computational and experimental reports.......................................................................................................................27 2-4 Comparison of the DFT geometric data for the vinylmethylene transition states and intermediates............................................................................................................28 2-5 Comparison of the DFT geometric data for the propenylidene transition state and intermediate..............................................................................................................29 2-6 Isodesmic of 1-fluorocyclopropene..........................................................................31 vii

PAGE 8

2-7 Comparison of fluorinated cyclopropenes...............................................................31 2-8 3-Fluorocyclopropene ring opening barriers............................................................37 2-9 Experimental and theoretical geometries of 3,3-difluorocyclopropene...................39 2-10 Experimental geometries of cyclopropene and 3,3-difluorocyclopropene..............40 2-11 B3LYP theoretical geometry of 3,3-difluorocyclopropene......................................41 3-1 Angles and dihedrals of the cyclopropenylcarbinyl ring opening............................57 viii

PAGE 9

LIST OF FIGURES Figure page 1-1 Structure of cyclopropane..........................................................................................5 1-2 Walsh orbital model...................................................................................................7 1-3 Two degenerate HOMOs of the refined Walsh..........................................................8 1-4 Frster, Coulson, and Moffitt’s cyclopropanes orbitals.............................................9 1-5 Molecular Orbitals #9, #11, #16, and #29 of cyclopropane.....................................10 1-6 The C 2v symmetric cyclopropene.............................................................................11 1-7 Walsh and “bent” models of cyclopropene..............................................................12 1-8 Resonance effects of fluorine on carbocations.........................................................15 1-9 Geometries of fluoromethyl radicals........................................................................16 1-10 Effects of and -fluorination on radicals.............................................................17 2-1 Proposed cyclopropene to methylacetylene routes..................................................20 2-2. The six possible vinylmethylene structures.............................................................21 2-3 Cyclopropene to vinylmethylene schematic............................................................24 2-4 The cyclopropene to propenylidene schematic........................................................24 2-5 Reaction mechanism for the allene to methylacetylene interconversions................25 2-6 Schematic of the potential energy surface and derived values.................................26 2-7 Cyclopropene-vinylmethylene ring opening............................................................28 2-8 Cyclopropene-propenylidene ring opening..............................................................29 2-9 Isodesmic of versus fluorine substitution.........................................................32 2-10 Geometries of the proximal vinylmethylene transition states..................................33 ix

PAGE 10

2-11 Vinylmethylene intermediates..................................................................................33 2-12 Distal bond breaking transition states......................................................................34 2-13 Ring opening schematic of the propenylidene intermediates...................................35 2-14 3-Fluorocyclopropene to propenylidene schematic.................................................36 2-15 3-Fluorocyclopropene’s vinylmethylene intermediates...........................................37 2-16 The C2v structure of 3,3-difluorocyclopropene.......................................................38 2-17 DFT HOMO’s and LUMO’s of cyclopropene.........................................................40 2-18 3,3-Difluoro vinylmethylene intermediate ring opening reactions..........................41 2-19 3,3-Difluorocyclopropene propenylidene reaction schematic..................................42 2-20 1,2-Difluorocyclopropene........................................................................................42 2-21 1,2-Difluorocyclopropene vinylmethylene ring opening.........................................43 2-22 The 1,2-difluorocyclopropene to propenylidene rearrangement..............................43 2-23 Schematic of tetramethylcyclopropene thermolysis................................................44 2-24 Schematic of 3-methyl and 3,3-dimethylcyclopropene thermolysis........................45 2-25 Schematic of 1,3,3-trimethylcyclopropene thermolysis...........................................45 2-26 Initial ring opening schematic of 1,3,3-trimethylcyclopropene...............................46 2-27 1-Fluoromethylcyclopropenes propenylidene formation schematic........................46 2-28 1-Fluoromethylcyclopropenes vinylmethylene formation schematic......................47 2-29 1-Fluoro-2-methylcyclopropene optimized structure...............................................48 2-30 1-Fluoro-2-methylcyclopropene to propenylidene intermediate schematic.............48 2-31 1-Fluoro-2-methylcyclopropene, Vinylmethylene intermediates............................49 2-32 1-Fluoro-3,3-dimethylcyclopropene........................................................................50 3-1 2,2-Difluorocyclopropylcarbinyl ring opening........................................................53 3-2 Cyclopropenylcarbinyl schematic............................................................................54 3-3 PES schematic of the cyclopropenylcarbinyl rearrangement...................................55 x

PAGE 11

3-4 Global minima of the C 4 H 5 radical..........................................................................56 3-5 Cyclopropenylcarbinyl Structures............................................................................57 3-6 The B3LYP PES schematic of the cyclopropenylcarbinyl rearrangement..............58 3-7 Cyclobutenyl cation structure...................................................................................58 3-8 HOMO of the 2-cyclobutenyl cation........................................................................59 3-9 The two initial 2-fluorocyclopropenylcarbinyl cation transition states...................59 3-10 Bond distances of 1-fluoro and 2-fluorocyclobutenyl cations.................................60 3-11 The 1-Fluoro-3-methylcyclopropenylcarbinyl minima............................................60 3-12 The HOMO of 2-fluorocyclobutenyl cation.............................................................61 3-13 The 6-31G(d) PES of the IRC calculations..............................................................62 3-14 The bisected transition state.....................................................................................63 3-15 The eclipsed and bisected transition states of the PCM calculations.......................64 3-16 PCM-IRC of the bisected and eclipsed transition states..........................................64 3-17 Cyclopropenyl Radical schematic............................................................................65 3-18 2-Fluorocyclopropenylcarbinyl radical, distal bond rearrangements.......................66 3-19 The endo-1-fluoro-bicyclobutanyl radical...............................................................67 3-20 The endo-exo transition state...................................................................................67 3-21 The exo-1-fluoro-bicyclobutanyl radical {3}...........................................................68 3-22 The proximal ring opening schematic......................................................................69 3-23 The endo-2-fluoro-bicyclobutenyl radical...............................................................70 3-24 The exo-1-fluoro-bicyclobutenyl radical {6}...........................................................70 3-25 SOMO of the bicyclobutenyl radical.......................................................................71 3-26 SOMO of the 1-fluoro-bicyclobutenyl radical.........................................................71 3-27 SOMO of the 2-fluoro-bicyclobutenyl radical.........................................................72 xi

PAGE 12

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy COMPUTATIONAL INVESTIGATION OF 1-FLUOROCYCLOPROPENES By G. Robert Shelton December 2004 Chair: William R. Dolbier Jr. Major Department: Chemistry The influence of fluorine on the thermal rearrangement of cyclopropenes has been investigated using computational methods. Also the investigation was expanded into studies of rearrangements of the 2-fluoro-3-methylcyclopropenyl radical and cation. Using Density Functional Theory (DFT) the systematic study of the initial ring openings of cyclopropene, 1-fluorocyclopropene, 3-fluorocyclopropene, 1,2-difluorocyclopropenes, and 3,3-difluorocyclopropenes. A novel and unprecedented “electrocyclic” automerization of 1-fluorocyclopropene was predicted. The automerization involves the preferred distal bond breaking and subsequent 180 rotation of its methylene group, without the formation of an intermediate carbene. The 1,2-difluorocyclopropene vinylmethylene intermediates were predicted to lie 4.93 and 8.79 kcal/mol higher in energy than the parent 1,2-difluorocyclopropene. In addition these intermediates have energy barriers of 16.8 and 24.9 kcal/mol to cyclize and reform 1,2-difluorocyclopropenes, making them good targets for trapping. xii

PAGE 13

1-Fluoro-3,3-dimethylcyclopropene was found to ring open preferentially via the distal bond with a ca. H = 21.06 kcal/mol into the transoid intermediate. The following 1,4-hydrogen migration was found to have a ca. H = 14.29 kcal/mol leading to the 2-fluoro-3-methyl-1,3-butadiene product. Finally the 2-fluorocyclopropenylcarbinyl radical and cation rearrangements were charted using DFT. The 2-fluorocyclopropenylcarbinyl cation was found to be a transition state leading to the aromatic global minima of 1-fluoro-3-methylcyclopropenyl cation, and 2-fluorocyclobutenyl cations. The fluorocyclobutenyl cations are shown to benefit from homoaromaticity in the case of the 2-fluorocyclobutenyl cation and fluorine stabilization of the positive charge in the case of the more stable 1-fluorocyclobutenyl cation. 2-fluorocyclopropenylcarbinyl radical was predicted to be a stable intermediate with a ca. H = 0.4 kcal/mol distal ring opening into the 2-fluoro-1,3-buten-1-yl radical. Using transition state theory, this reaction was calculated to have a rate of 7.68x10 12 s -1 . Some five orders of magnitude faster than the proximal ring opening, and an order of magnitude faster than the previously studied 2,2-difluorocyclopropylcarbinyl radical opening. xiii

PAGE 14

CHAPTER 1 FLUORINE SUBSTITUENT EFFECTS Introduction There are three distinctive characteristics that make the fluorine atom unique as a substituent. Fluorine has the highest electronegativity of any element and therefore has an ability to polarize its bond to carbon. The atom has three non-bonded electron pairs in 2p orbitals of similar size and energy to carbon, allowing for the overlap of the 2s and 2p orbitals of carbon. And finally fluorine has a relatively small size, thus minimizing steric interactions. These three characteristics all combine to cause a unique substituent effect when fluorine is incorporated into organic molecules and reactive intermediates. These substituent effects when understood for a particular system can make for an interesting set of tools for use in structure and reactivity studies. The Pauling scale is a relative scale based on the tendency of an atom directly involved in a bond to attract the bonding electrons. This polarity and ionic nature of a bond can be gauged by the difference between the electronegativity of the two atoms concerned. Table 1-1 shows the electronegativity values 1 of several common atoms found in organic systems are listed. As is most important in organofluorine chemistry the bond of interest is that of carbon and fluorine. The difference between these atoms’ electronegativity in particular results in a strong inductive withdrawing effect of the electron density through the molecular framework. The resulting bond is a highly polarized bond with substantial ionic character. 1

PAGE 15

2 Table 1-1. Electronegativity of several common atoms Atom Electronegativity Atom Electronegativity Hydrogen 2.20 Phosphorous 2.19 Carbon 2.55 Sulfur 2.58 Nitrogen 3.04 Chlorine 3.16 Oxygen 3.44 Bromine 2.96 Fluorine 3.98 Iodine 2.66 The three non-bonding pairs of electrons in the 2p orbitals of fluorine are of similar size and energy to those of the other second row elements. When the geometry of the molecule allows for overlap of the p-orbitals with electron deficient sites the potential for strong electron donation can occur. This resonance allows for an offsetting back donation of electron density from fluorine into a molecular site, in contrast to the -withdrawing nature of this most electronegative atom. To understand the relative small size of fluorine, one must consider the van der Waals radius 1 of fluorine relative to that of the other halogens, as found in Table 1-2 below. Its small radius allows fluorine to be the least sterically demanding substituent, especially among the halogens. Table 1-2. Van der Waals Radii Atom Radius () Atom Radius () Fluorine 1.47 Carbon 1.70 Chlorine 1.73 Oxygen 1.52 Bromine 1.84 Hydrogen 1.20 Iodine 2 2.01 The above dichotomy of fluorine’s electronic effect (-withdrawing, -donating), coupled with its relatively small size, allows for the majority of fluorine’s influence on reactivity and structure to arise from its electronic influence on the system. However this dichotomy also make the prediction of the exact substituent effect difficult.

PAGE 16

3 Structure, Bonding and Reactivity Saturated Systems Table 1-3 gives the carbon-halogen interatomic distances for halomethanes and ethanes. 2, 3 The unique trend observed is a shortening of the C-F bond in conjunction with increasing bond dissociation energy (BDE). The same trend is observed in the geminal substituted ethanes. However the trend is broken once vicinal fluorination is examined. The carbon-carbon bond distance of CH2FCF3 is 1.501, with a BDE of 94.6 kcal•mol -1 compared to that of CF3-CF3, which increases both the distance and BDE to 1.545 and 98 kcal•mol -1 . Table 1-3. Carbon-halogen bond lengths () over the BDE (kcal•mol -1 ) X CH3X CH2X2 CHX3 CX4 CH3-CH2X CH3-CHX2 CH3-CX3 F 1.385 108.3 1.357 119.5 1.332 127.5 1.319 130.5 1.398 107.9 1.343 Unknown 1.335 124.8 Cl 1.781 82.9 1.772 81.0 1.758 77.7 1.767 72.9 1.802 1.790 1.771 Br 1.939 69.6 1.934 64.0 1.930 62.0 1.942 56.2 To rationalize the above trends, a hypothesis based on hybridization was put forth. It is proposed that the difference in electronegativity between carbon, hydrogen, and fluorine causes rehybridization to occur increasing the amount of p character of the carbon fluorine bond. Thus the C-F bond possesses greater p character, with greater s character being found in adjacent C-H bonds. This rationale not only accounts for the incremental C-F bond strengthening, but also for the observed changes in the geometry within the fluoromethane series. Accumulation of p character in C-F bonds should lead to a decrease in the F-C-F bond angle. This is consistent with the experimental observation outlined in Table 1-4. 4, 5

PAGE 17

4 Table 1-4. Experimental fluoromethane angles Methane F-C-F H-C-H CF4 109.5 -CHF3 108.7 -CH2F2 108.3 113.7 Unsaturated Systems Vinylic fluorine substitution results in shortening of C=C bonds compared to the parent hydrocarbon, as well as shorter C-F bonds than in similar fluoroalkanes (Table 1-5) 2, 3 . Ab initio calculations indicate the C-F shortening to be caused by delocalization (-donation) of fluorine’s 2p electrons into the C=C bond. 6 Table 1-5. Bond distances, angles, and BDE of fluoroethenes Parameter CH2=CH2 CH2=CHF CH2=CF2 CHF=CF2 CF2=CF2 r (C=C), 1.339 1.333 1.316 1.309 1.311 r (C=C), -1.348 1.324 1.336 1.319 HCH 117.8 114.7 119.3 --HCF -111.3 -114.0 -FCF --109.7 109.1 112.6 BDE, kcal•mol-1 63-64 Unknown 62.8 Unknown 52.3 The heats of hydrogenation (H) for the series of fluoroethenes are provided in Table 1-6. 1 The lowering H trend illustrates the increase in reactivity of fluorinated alkenes. In general, hydrogenation of a fluorinated olefin into the saturated derivative is an exothermic process compared to that for the parent hydrocarbon. This is attributed to the combination of the withdrawing effect of polyfluorination on the framework and the thermodynamic preference for geminal difluorine substitution on saturated carbon centers. Table 1-6. Heats of Hydrogenation Ethene H (H2), kcal•mol-1 CH2=CHF -29.7 CH2=CF2 -38.8 CF2=CF2 -45.7

PAGE 18

5 Cyclopropane Structure In 1946 the first gas phase electron diffraction study of cyclopropane was published 7 and since then reaffirmed by gas phase microwave spectroscopy and crystal X-ray diffraction. Figure 1-1 shows cyclopropane’s three member ring structure with D 3h symmetry. The agreements among experimental and computational methods are quite good and are outlined in Table 1-7. 8 Figure 1-1. Structure of cyclopropane Table 1-7. Structure parameters of cyclopropane C-C () C-H () H-C-H (deg) Experimental 9 1.512 1.083 114.0 HF 10 1.497 1.076 114.2 MP 11 1.502 1.084 114.2 B3LYP 12 1.508 1.087 113.9 It is interesting to note the computationally predicted C-C bond distances 12 in cycloalkanes (Table 1-8). The C-C bond length has a maximum value in cyclobutane and minimum length in cyclopropane. Replacement of any one of the methylene groups in cyclopropane by either an oxygen or a nitrogen results in even shorter C-C bond distances of 1.466 and 1.481 respectfully. The C-H bonds are the shortest in cyclopropane at 1.084 compared to cyclobutanes 1.093 or even benzenes 1.083. This is an indication of cyclopropane’s high s character in that particular bond.

PAGE 19

6 Table 1-8. Bond lengths in different cycloalkanes 12 C3H6 C4H8 C5H10 C-C () 1.497 1.549 1.534(1) C-H () 1.084 1.093 Strain Baeyer 13 first introduced the concept of ring strain in 1885. In modern organic chemistry, strain effects are generally discussed in terms of bond angles, lengths, and torsion distortions. 14 The strain energy of a molecule is the additional energy of a system when compared to a suitable strain free system. Cyclohexane is usually considered to be the model of a cyclic strain free system. The cylcohexane molecule has a heat of formation (H f ) of -29.5 kcal/mol which averages out to -4.92 kcal/mol/methylene. The fact that a methylene unit in the straight chain counterpart, hexane, has a value of -4.926 kcal/mol makes this a very good starting point even with the gauche interactions that occur in cyclohexane. Using this “group equivalent” basis for cyclopropane the ‘strain-free’ energy works out to be -14.76 kcal/mol. When compared to the heat of formation, +12.73 kcal/mol, the difference is considered to be the strain energy of the molecule, 27.5 kcal/mol! Carrying out the same logic on the strain energies of cyclobutane and cycloheptane, the numbers work out to be 26.5 and 6.2 kcal/mol respectively. 12 Several computational methods are now accepted means of determining heats of formation. Calculation performed at B3LYP/6-31G(d)\\HF/6-31G(d) with a Zero Point Energy (ZPE) correction led to a calculated ring strain 12 of 28.2 kcal/mol for cyclopropane. Electronic structure Walsh 15 first proposed an orbital model for the bonding in cyclopropane in 1949. In chosing sp 2 hybridization for the carbon, it allowed for one hybrid to be used for each C

PAGE 20

7 H bond and the third pointing directly into the center of the ring (figure 1-2). The unhybridized p orbital was then shown to lye in the plane of the ring. The three sp 2 hybrids that pointed toward the ring center can overlaped and form what would be three molecular orbitals (MO), one bonding and two antibonding. This mixing between the end-on type characteristic of bonds and the side-on type characteristic of bonds goes to form the three delocalized orbitals, generating a total of three (two bonding and one antibonding) orbitals. Figure 1-2. Walsh orbital model The Walsh orbitals attracted much attention since this model seemed to be close to the canonical SCF MOs of cyclopropane and it further attempts to explain many of the unusual aspects of cyclopropane chemistry. These included the character of its C-C bonds, substituent effects on bonding, structure, and strain, not to mention the ability to delocalize electrons in the plane of the ring. Walsh orbitals however suffered from a number of draw backs that were constantly refined to correctly describe the properties of the cyclopropyl group. 8 Refined Walsh orbitals were obtained by assuming a set of sp hybrid orbitals complemented by two p orbitals for each C atom. These orbitals were radially oriented to the center of the ring (sp in and sp out orbitals) and tangentially oriented to the perimeter of the ring (p (in-plane) and p (out-plane) orbitals) as depicted in Figure 1-3. The six Walsh orbitals

PAGE 21

8 are formed from the radial sp in and the tangent p (in-plane) orbitals. These mixed orbitals can be classified as a ‘-bridgedorbital’ (-orbital bridge at C1) and a ‘bridged--orbital’ (-orbital bridged at C1). Radial o r ientationTangento r ientation Figure 1-3. Two degenerate HOMOs of the refined Walsh Moving away from the Walsh orbitals as it became more apparent that this was not a correct description of the ground state of cyclopropane, a bent bond model of cyclopropane was introduced by Frster 16 , and later refined by Coulson and Moffitt 17 . Ingold actually proposed the idea of bent bonds in 1920, but with questionable experimental evidence. In all they attempted to describe the hybrid orbitals with optimized sp 3 hybridized carbons, which lead to the orbital picture depicted in Figure 1-4. Together they describe increased p-character of the CC hybrid orbitals via sp 4 rather than sp 2 character for the bond. While the the s-character of the sp m hybrid orbitals was increased from sp 3 to sp 2 . The Frster-Coulson-Moffitt further demonstrates that the sp n orbitals are considerably bent, resembling the strain-free tetrahedral angle. Therefore the Frster-Coulson-Moffitt orbitals demand a weakening of the CC bonds, and thereby straining the three-membered ring.

PAGE 22

9 Figure 1-4. Frster, Coulson, and Moffitt’s cyclopropanes orbitals Fluorine substituent effect In 1966 the first report 18 appeared documenting the kinetic substituent effect fluorine has on cyclopropane. It was observed that an increase of 4.5~5 kcal/mol/fluorine of stain was introduced into the cyclopropane ring. It was not until Rchardt in 1997 reported enthalpy of combustion data for strain-free fluorinated group equivalents in combination with Roth’s heats of hydrogenation from five years earlier allowed the experimental assignment of 41.8 kcal/mol for the strain of a 1,1-difluorocyclopropane ring. Carrying this out to the incremental strain due to the geminal fluorine substituents is 14.2 kcal/mol. Wiberg’s computational studies focused around a isodesmic reaction provided similar results of 11.2 to 13.3 kcal/mol of incremental strain due to the geminal fluorine. Dolbier later, using 1,1,4,4-tetrafluorocyclohexane instead of Wiberg’s 2,2-difluoropropane, was able to calculate a strain energy of 37.2 kcal/mol for 1,1-difluorocyclopropane. When compared to the ca. 28.2 kcal/mol for cyclopropane, an incremental strain due to the geminal fluorine of 9 kcal/mol/fluorine is deduced. In today’s literature a ring strain increase of about 5-7 kcal/mol/fluorine for cyclopropane is widely accepted.

PAGE 23

10 Both structural and theoretical studies indicate that fluorine substituents on a cyclopropane ring shorten the proximal C-C bonds and lengthen the distal bonds. Microwave spectroscopy found the FCF in 1,1-difluorocyclopropane, 108.4, to be much smaller than the 114.5 HCH in cyclopropane. Fluorine’s non-bonding electron pairs can interact with a conjugated system through conjugative interaction to either destabilize a filled system as discussed earlier. It may also use this -donation to stabilize a vacant system as in the case of carbocations, which will be discussed next. The only requirements for these interactions are comparable orbital energies for the substituent as well as the ring. An extensive MO search by Clark, Schleyer and coworkers have found that orbitals #9, #11, #16, and #29 of cyclopropane would allow for orbital interactions of fluorine and carbon (Figure 1-5). In particular and most striking is the high s character of MO #9. When substituted with the -withdrawing fluorine substituent it should be easily seen how this would result in the lengthening of the distal and shortening of the proximal C-C bonds. CCCH2 #16CC #29CC #11CH2 #9 Figure 1-5. Molecular Orbitals #9, #11, #16, and #29 of cyclopropane

PAGE 24

11 Cycopropene Structure Microwave, electron diffraction, NMR, and X-ray diffraction have been combined to resolve the structural parameters of cyclopropene and many of its derivatives. This C 2v symmetric molecule (Figure 1-6) has C-C bond lengths remarkably similar of cyclopropane (Table 1-9). The very short C=C bond when compared to ethylene (0.04 smaller) and allene(0.01 shorter) are of particular interest. When comparing the similar situation of the cyclopropane C-C bond to the unstrained counterpart a difference of 0.025 is seen. It is deduced that the shorting in the double bond of cyclopropene comes from increased s-character in the network coupled with strong p overlaps. Figure 1-6. The C 2v symmetric cyclopropene The methylene fragment of the cyclopropene has an angle and bond length to back up a vinylic description of those particular protons. The true vinylic protons are rather short themselves and this implies that a greater s-character must be seen in the C=C bond. Table 1-9. Geometric data for cyclopropene Bonds and angles Microwave spectra 19 Computational C-C () 1.509(1) 1.509 C=C() 1.2959(4) 1.295 C-H (methylene, ) 1.088(2) 1.095 C-H (vinyl, ) 1.072(1) 1.080 C=CH() 149.85(8) 149.896 HCH() 114.5(2) 112.995 C 1 C 3 C 2 () 50.84(5) 50.824 C 1 C 2 C 3 () 64.58 64.587

PAGE 25

12 Strain Whereas cyclopropane has a calculated ring strain of 28.2 kcal/mol, cyclopropene has a calculated value of 54.5 kcal/mol. The molecular geometry of cyclopropene heavily suggests this additional strain is held within the framework. Geometry observations suggest this in large part via the relatively small C 1 C 2 C 3 and C 1 C 3 C 2 angles. Electronic structure Building off the two descriptions outlined in the cyclopropane section, you would get the following diagrams for the Walsh and Frster, Coulson,Moffitt’s ( or rather ”Bent”) description (Figure 1-7). In the extrapolated Walsh model the vinylic carbons are sp-hybridized, one p-orbital on each is then used in double bond formation while the remaining contributes to the ring bonds. The methylene carbon is sp 2 hybridized as it was in the cyclopropane model. In the bent model the framework is now formed from four, two from each vinylic carbon, sp 3 hybrids and two sp 5 hybrids from the methylene carbon. The vinylic carbon orbitals that are used to form the bonds to substituents in cyclopropene are now sp in character. Similar to that of acetylene. The methylene carbon remains sp 2 in character for substituent bonding. Figure 1-7. Walsh and “bent” models of cyclopropene

PAGE 26

13 Fluorine substituent effect The only fluorinated cyclopropenes synthesized are 3-fluorocyclopropene, 3,3-difluorocyclopropene, 1,3,3-trifluorocyclopropene and the perfluorinated 1,2,3,3-tetrafluorocyclopropene. Of the four molecules 3,3-difluoro and the perfluorinated cyclopropenes are the most studied experimentally and theoretically. And from the author’s personal experience, the attempted synthesis of 1-fluorocyclopropene will haunt him until accomplished. Wiberg has shown though a series of isodesmic equations that fluorine prefers to be attached to a carbon bearing other carbons rather than hydrogens (Table 1-10, equation 1). This should not be surprising when considering the discussion earlier with the fluorine substituent effect with saturated systems. The addition of methyl groups to the electronegative C-F bond allows the methyl group to donate electron density through the framework there by stabilizing the increased positive charge on the C of the C-F bond. Table 1-10. Wiberg’s isodesmic equations and computed results H (kcal/mol) Isodesmic Reaction MP2 MP2* B3LYPCH4[1]+F2C CH2F2+H2C 20.6 18.316.6 [2]+F2C +H2C F 2C 13.3 14.211.2[3]+F2C +H2C F 2C 3.6 4.21.9[4]+ + F 2C F 2C -9.7 -10.0-9.3 The series also indicates (equation 2) that fluorine prefers to be attached to propane rather than cyclopropane. This further enhances the discussion of the substituent effect on saturated system. That is due to C orbitals in propane being sp 3 hybridized but cyclopropane uses sp 2 hybridized orbitals. Showing that fluorine prefers carbons with

PAGE 27

14 higher p character. Equation 3 also supports this idea and further the enhancement of vinylic character in the methylene carbon of cyclopropene. Equation 4 however shows the unpredictability of the fluorine substituent effect. From this approximate ca. H =-10.0 kcal/mol it seems as if the difluoro substitution stabilizes the cyclopropene. Further isodesmics have indicated that =0 leads to a similar but larger stabilization. This has been dismissed as being more to do with polarization of the carbonyl group leading to cyclopropenium (aromatic) ion character of the cyclopropene. Perfluorocyclopropene has been characterized by microwave spectroscopy, electron diffraction, liquid-crystal NMR, and X-ray diffraction. 3,3-difluorocyclopropene’s structure has been determined microwave spectroscopy as well. Table 1-11 compares these two cyclopropenes with the hydrocarbon parent, although one must be careful drawing conclusions from data using different methods. There is however a clear indication of considerable shortening of the single bonds within the cyclopropene, consistent with the explanations given for cyclopropane. Table 1-11. CC bond lengths(r/pm) for several cyclopropenes Method r(C=C) R(C-C) Cyclopropene ED 130.4(3) 151.9(12) 3,3-difluoro MW 132.1(1) 143.8(7) ED/MW/LC 130.7(13) 146.1(3) Perfluoro XRD 129.6(4) 145.3(3) Reactive Intermediates Carbocations Fluorine’s -donating ability to the vacant p orbitals of a cation dominates it’s would be destabilizing -withdrawing ability. This -donating resonance effect is

PAGE 28

15 responsible for the stabilization of a carbocationic center when fluorine is alpha substituted. On the contrary, when beta substitution occurs the -withdrawing inductively destabilizes the cation by placing a partial positive charge on the carbon adjacent to the carbocation (Figure 1-8). C F C F StabilizationC C F Destabilization Figure 1-8. Resonance effects of fluorine on carbocations These contrasting effects go to explain the chemistry seen in electrophilic reactions 20 . Perfluoroalkenes are generally unreactive towards electrophilies whereas fluoroalkenes usually react to minimize the beta substitution. And in the case were the positive charge develops on the alpha substituted carbon, the transition state must be very late. Free Radicals The unpaired electron in a radical is in a semi occupied molecular orbital (SOMO), a somewhat similar situation to that of the completely vacant orbital in a carbocation species. And as discussed previously the -withdrawing character of the fluorine substituent will interact with the SOMO via the C-F * orbital. This effect manifests itself in the structural deviations from the parent hydrocarbon. But more importantly this should lower the SOMO energies and therefore should show enhanced SOMO-HOMO interactions. This leads to changes in hydrogen abstraction and addition reaction reactivity.

PAGE 29

16 Successive fluorination directly upon the methyl radical causes the normal structurally planar, -type radical to become increasingly pyramidal and the radical center to develop more -type character. This has been confirmed experimentally and computationally by ab initio and paramagnetic resonance measurements. Calculations especially show the increase of inversion barriers for increasing fluorinated methyl radicals, and the effect on geometries (figure 1-9). Figure 1-9. Geometries of fluoromethyl radicals The effect of fluorine on stabilization is progressively decreasing with -fluorination. This was best described by the following isodesmic equation (table 1-12) which was used to describe the radical stabilization energies (RSE). Table 1-12 also leads the discussion into -fluorination. The effect on RSE is essentially the same for -fluorination. The calculated RSE for the monofluoromethyl radical is greater than both difluoroand the much destabilized trifluoromethyl radical. Fluorine’s effect on radical

PAGE 30

17 stabilization is rather minuscule compared to that of a methyl substituent, RSE’s for one methyl and three methyl groups which are 3.3 and 8.0 kcal/mol, respectively Table 1-12. Radical stabilization energies (RSE) as described by the isodesmic equation Xn•CH3-n + CH4 XnCH4-n + •CH3 X RSE (Kcal/mol) X1 RSE (Kcal/mol) F +1.64 CH3 +3.27 F2 +0.56 CH2F +1.46 F3 -4.21 CHF2 +0.16 CF3 -1.34 Fluorine occurring adjacent to a radical center has also been probed theoretically for the effects on radical geometry. It was found however that -fluorination has relatively no effect on the radical center (figure 1-10). It is worthy to note that -fluorination has approximately the same effect on ethyl radicals as it does on methyl radicals. Figure 1-10. Effects of and -fluorination on radicals Carbene For the greater part of this dissertation only the singlet state of carbenes will be considered. And before now the -withdrawing effects have played the major role in the

PAGE 31

18 substituent effects. Now the donating ability of fluorine when adjacent to a carbene goes to stabilize the singlet state compared to the triplet. It has been shown experimentally and computationally that the singlet-triplet energy gap of HCF is approximately 14.7 Kcal/mol. This thermodynamic stabilization is quite interesting when you consider the kinetic stabilization caused by -substitution. The -withdrawing behavior of fluorine inhibits the electron deficient C-F bond to undergo 1,2 fluorine migration. This lack of alkene formation is the cause of the kinetic stabilization. Effort spent on structural determination of fluorocarbenes has focused on rotational analysis. It has been determined that ground state singlet structure C-F bond length is 1.31 and FCH is 101.8 to 104.1. Its interesting to note that the triplet state shows little change in bond lengths but the FCH opens up to 123.8.

PAGE 32

CHAPTER 2 THEORETICAL STUDY OF THE THERMOLYSIS OF 1-FLUOROCYCLOPROPENE Part 1: The C 3 H 4 & C 3 H 4 F Initial Ring Opening Introduction Thermolysis of cyclopropene 21 at 495K results in the unimolecular isomerization into the major product methylacetylene (55.2%), minor product allene (2.4%), and the remaining percentage was unchanged cyclopropene. Early thermolysis studies 22 of the allene to methylacetylene conversion were undertaken with cyclopropene as the assumed intermediate. Furthermore, deuterium labeling results 23 suggest that 50-100% of the methylacetylene is formed through the cyclopropene intermediate. Shock tube experiments 24, 25 also confirmed the intermediacy of cyclopropene. From the earliest studies of Srinivasan, 21 propene-1,3-diyl was the proposed intermediate for the cyclopropene to methylacetylene rearrangement. 26, 27 This diradical would then undergo a 2,3 H-shift producing the methylacetylene product (Figure 2-1). This proposed pathway was seemingly confirmed by Walsh’s 21, 28, 29 work and separately by Bergman’s 30 work with optically active cyclopropenes. However, Bergman found evidence for and proposed a two step, carbene type intermediate (vinylmethylene) in place of the 1,3-diyl for the cyclopropene isomerization. This carbene intermediate was nothing more than a resonance hybrid of the given 1,3-diyl that would benefit from allylic stabilization. 19

PAGE 33

20 H2CHCCH CH2CCH2 H2CC CH2 H2CHC C H H3CC C H H3CC CH CHC CH2 H (1,3-dyl)propenylidene(vinylmethylene) Figure 2-1. Proposed cyclopropene to methylacetylene routes. To completely round out the discussion two more possible intermediates must be introduced in the thermal rearrangement of cyclopropene to allene to methylacetylene. A 1,2-hydrogen shift from C1 to C2 of cyclopropene would lead to the cyclopropylidene intermediate (top left of Figure 2-1). A sequential 1,2-hydrogen shift along C1 and C3 with ring opening leads to the propenylidene intermediate (bottom of Figure 2-1). The propenylidene intermediate as it turns out is the favored intermediate of the cyclopropene to methylacetylene rearrangement. The vinylmethylene intermediate is the favored intermediate cyclopropene to allene conversion. Starting from cyclopropene the barrier to cyclopropylidene 31 is 15.5 kcal/mol higher than the ring opening into the transoid and 21.1 kcal/mol higher than the cisoid vinylmethylene. Starting from allene, the cyclopropylidene intermediate was found to have an energy barrier 5.4 kcal/mol 31 higher than the barrier to vinylmethylene. Thus this intermediate was not considered as a possibility in this computational study.

PAGE 34

21 The C 3 H 4 Surface: Ab Initio Study Building off the proposed vinylmethylene intermediates of Walsh and Bergman, Yoshimine, et al published a series of papers, and their computational model has become the accepted mechanism for the ring opening of cyclopropene. Structures and stabilities of isomers Six possible vinylmethylene intermediates were proposed when considering the early diradical and carbene intermediates (Figure 2-2). In Yoshimine’s first study self-consistent-field (SCF) wavefunctions with a doubleplus polarized (DZP) functional basis set was used for their computations. However, It was found that the cisand transforms of the allylic structure were isoenergetic and had a ground state higher in energy than methylacetylene. No other planar or bisected structures were found to be stable points on the singlet surface. C1C2 C3 H H H H trans planar carbeneTPC, (1A') C1C2 C3 H H H H cis planar carbeneCPC, ( 3,1A" ) C1C2 C3 H H H H trans planar diradical TPD (3,1A") C1C2 C3 H H H H cis planar diradical CPD (3,1A") C1C2 C3 H H H H trans bisected diradical TBD (3,1A')cis bisected diradical CBD (3,1A')C1C2 C3 H H H H Figure 2-2. The six possible vinylmethylene structures.

PAGE 35

22 Earlier computational 32, 33 results suggested that some of the possible carbene and diradical intermediates may have had the ground state electronic structure of a triplet. An extensive examination of the propenylidene geometry (Figure 2-1) and electronic states was performed by Yoshimine, and it demonstrated that the singlet state is 41.0 kcal/mol lower in energy than the triplet state. Of the six vinylmethylene geometries (Figure 2-2), four triplet and six singlet states were found to be minima. Triplet states of TPC and CPC were the most stable, followed by their respective singlets, but only by 9.7 and 11.5 kcal/mol. ESR experiments suggested a triplet-singlet energy separation of less than 1 kcal/mol for TPC and CPC. 34 . It was proposed by Yoshimine that a method incorporating “resonance energy” could account for the disparity of the experimental results, and this would possibly lead to the computationally predicted singlet ground state as the most stable. According to the conclusion of the ab initio study by Yoshimine, MCSCF and CI calculations were employed to account for the electron correlation lacking in the SCF(DZP) method. It was found that the bisected 1,3-diradicals were not stable intermediates, but the allylic singlet and triplet states were stable. This study also showed that the singlet TPC and CPC are lower in energy than their triplet states, requiring the crossing of a 5.7 kcal/mol barrier. With better electronic descriptors available for the energy, the geometries at SCF (4-31G) of the three stable ground states are listed in Table 2-1. These geometries agree excellently with the experimentally determined values. Table 2-2 outlines the geometries of the reactive intermediates found lying on the singlet surface of the allene, cyclopropene, methylacetylene rearrangement.

PAGE 36

23 Table 2-1. Computed and experimental geometries of the most stable C 3 H 4 isomers. Methylacetylene Allene Cyclopropene SCF(DZP) Exp. 35 SCF(DZP) Exp. 36 SCF(DZP) Exp. 19 C1C2 1.472 1.459 1.302 1.308 1.500 1.509 C2C3 1.191 1.206 --1.281 1.296 C1H1 1.084 1.105 1.077 1.087 1.084 1.088 C2H3 ----1.070 1.072 C3H4 1.060 1.056 ----C1C2C3 ----64.7 64.6 H1C1C2 110.2 110.3 121.0 120.9 119.7 119.1 H3C2C3 ----150.1 149.9 H1C1H2 ----113.6 114.6 Table 2-2. Geometries of intermediates on the singlet surface. Vinylmethylene PropenylideneTrans Cis C1C2 1.510 1.341 1.340 C2C3 1.298 1.453 1.461 C1H1 1.084 1.076 1.078 C1H2 1.079 1.079 C1H3 1.083 C2H3 1.081 1.079 C2H4 1.085 C3H4 1.097 1.097 C1C2C3 127.2 117.1 123.6 H1C1C2 110.1 119.9 121.6 H2C1C2 122.4 121.7 H3C1C2 111.1 H3C2C3 124.3 117.8 H4C2C3 112.1 109.2 106.9 Thermal isomerization The resulting surface from the SCF and MCSCF studies showed the lowest energy pathway for cyclopropene’s conversion into methylacetylene would likely occur either through a propenylidene or vinylmethylene intermediate. It was found that the cyclopropene to vinylmethylene conversion is highly reversible with an energy barrier of 32.9 kcal/mol required to open into the cisoid vinylmethylene and 35.8 kcal/mol to open into the transoid intermediated. As mentioned before, these are allylic-like structures with

PAGE 37

24 the terminal hydrogen slightly above or below the plane of the carbon backbone. A 1.3 to 4.7 kcal/mol is required to go back to the cyclopropene from the cisoid and transoid form respectively (Figure 2-3). C3C2C1HHHHC3C2C1HHHHC3C2C1HHHH35.84.532.91.1TransCisEnergy barriers are in kcal/mol and include zero-point corrections. Figure 2-3. Cyclopropene to vinylmethylene schematic. The cyclopropene to propenylidene reaction surface involves a 1,2-H shift with simultaneous electronic rearrangement to form propenylidene. The potential energy barrier for this reaction is 38.1 kcal/mol. This step is not as reversible as the vinylmethylene conversion, because it requires 18.6 kcal/mol for the propenylidene to cyclopropene isomerization (Figure 2-4). SCF (DZP) examination of the propenylidene intermediate shows that in actuality at this level of theory it is not a stable isomer, due to a barrier free 1,2-H shift leading to methylacetylene. C3C2C1 H H H H C3C2 C1 H H C3C2 C1 HH H H 41.518.6Energy barriers are in kcal/mol and include zero-point corrections. H H Figure 2-4. The cyclopropene to propenylidene schematic. The 1,2-H shift required for transoid vinylmethylene to form allene has an energy barrier of 29.5 kcal/mol. However, if the cyclopropene had opened up into the lower energy cisoid form, the energy barrier is only 12.6 kcal/mol to form allene.

PAGE 38

25 In concluding the cyclopropene to methylacetylene or allene reaction mechanism, the kinetic ring opening of cyclopropene favors the formation of the cisoid vinylmethylene intermediate. However, another energy barrier of almost 30 kcal/mol must be overcome to form allene. The next lowest energy barrier for ring opening is that which forms the transoid vinylmethylene intermediate. A 12.5 kcal/mol barrier must be over come for this transformation. The overall pathway has a rate determining step dependent upon the transoid vinylmethylene’s 1,2-H shift barrier in forming the product. The highest initial energy barrier from cyclopropene forming propylidene results in the thermodynamic and kinetically favored methylacetylene product (Figure 2-5). Figure 2-5. Reaction mechanism for the allene to methylacetylene interconversions Computational detail All structures were optimized using density functional theory (DFT) calculations within the Gaussian 98 and 03 program packages. The level of theory was set to Becke’s hybrid three-parameter functional (B3LYP), and using the 6-31G(d) basis set, frequency calculations were performed on all stationary points to identify transition structures and to determine thermochemical information. Transition structures were characterized by a single imaginary frequency. Thermal energies and entropy terms were obtained using

PAGE 39

26 frequencies scaled by 0.9807 at 298.15 K and 1 atm. An intrinsic reaction coordinate (IRC) calculation was performed for each transition structure to follow the entire reaction pathway for each transition structure. Single-point energies were calculated using B3LYP level of theory and the 6-311+G(2df,2p) basis set. Calculated rate constants were derived from transition state theory. Figure 2-6. Schematic of the potential energy surface and derived values. The reported values used through out are depicted in Figure 2-6. Activation barriers (E 0 ) refers to the difference between the zero point corrected electronic energies of transition state and reactant. The reaction energy (E rxn ) refers to the zero point corrected electronic energy difference between product and reactant. Standard enthalpy of activation (H ) is obtained after the thermal correction of E 0 to 298.15 o C, 1 atm. The heat of reaction (H) refers to the difference of enthalpies between the reactant and

PAGE 40

27 product of a reaction at 298.15 K and 1 atm. which was obtained after the thermal correction . The C 3 H 4 Surface: DFT study The main focus of this study was the initial ring opening of cyclopropene and how ultimately fluorine might influence these initial steps through geometric and electronic changes. It was first prudent to demonstrate the utility and accuracy of DFT calculations on cyclopropene rearrangements, and as a DFT basis for comparison of the fluorinated analogs. Structures and stabilities of isomers Cyclopropene was optimized to the C 2v point group, and the energy calculated at the previously mentioned level of theory. The resulting geometric parameters are laid out in Table 2-4. All three are in good agreement to one another. Table 2-3. DFT geometric data and comparison to previous computational 31, 37-39 and experimental 19 reports. Parameter B3LYP 6-31G(d) SCF(DZP) Exp. r(C-C) () 1.509 1.500 1.509 r(C=C) () 1.295 1.281 1.296 r(=C-H) () 1.080 1.084 1.088 r(C-H) () 1.095 1.070 1.072 (C-C(H2)-C) 64.589 64.7 64.6 (=C-C-H) 119.905 119.7 119.1 (C=C-H) 149.894 150.1 149.9 (H-C-H) 112.995 113.6 114.6 The lowest energy barrier for ring opening was previously calculated to be the transition state leading to the cisoid and transoid vinylmethylene intermediates. This reaction is depicted in Figure2-7. From the geometry of the transition states, with respect to the methylene rotation, it can be seen how late they must occur in the reaction. The geometric parameters are presented in Table 2-4. The numbering scheme assumes the

PAGE 41

28 C1-C3 bond is broken. Neither of the transition states nor vinylmethylene products had symmetry higher than C 1 . The cisoid and transoid planar carbenes are transition states that connect the respected ground state structures. Figure 2-7. Cyclopropene-vinylmethylene ring opening. Table 2-4. Comparison of the DFT geometric data for the vinylmethylene transition states and intermediates. B3LYP 6-31G Parameter VC VC CP VT VT r(C1=C2) () 1.390 1.373 1.295 1.383 1.415 r(C2-C3) () 1.381 1.391 1.509 1.380 1.365 (C1C2C3) 109.7 104.8 64.6 101.7 111.6 (C1C2C3H6) 36.3 45.6 105.9 32.3 16.0 (H5C1C2C3) 38.3 48.3 180.0 138.6 137.8 MCSCF Parameter VC VC CP VT VT r(C1=C2) () 1.378 1.338 1.281 1.336 1.401 r(C2-C3) () 1.397 1.467 1.500 1.509 1.403 (C3C1C2) 122.9 122.2 64.7 115.0 121.2 (C1C2C3H6) 9.1 6.8 105.6 37.2 1.7 (H5C1C2C3) 41.9 27.1 180.0 164.6 146.1 Yoshimine and coworkers found this highly reversible reaction to have E 0 of 32.6 and 35.8 kcal/mol for the cyclopropene to cisoid and transoid vinylmethylene ring opening. The B3LYP level predicts the same E 0 to be 35.2 and 35.4 kcal/mol for the cisoid and transoid vinylmethylene conversion. However the reverse was previously reported to have an activation energy of 4.7 kcal/mol for the transoid and 1.3 kcal/mol for the cisoid. In this study the reverse E 0 and H of the cisoid was negative indicating a barrierless transition state for cyclization. This is a clear indication of the very late

PAGE 42

29 transition state and the very unstable cisoid intermediate that was calculated by B3LYP. The transoid intermediate was slightly more stable, with respect to the cisoid, with a ca. E 0 =0.08 kcal/mol but once the thermal correction was made the barrier was nonexistent. What turned out to be the kinetic and thermodynamic pathway in cyclopropene thermal isomerization was the ring opening and simultaneous 1,2 H-shift that lead to the propenylidene intermediate (Figure 2-8). Table 2-5 compares the geometries of this study to the Yoshimine study. Figure 2-8. Cyclopropene-propenylidene ring opening. Table 2-5. Comparison of the DFT geometric data for the propenylidene transition state and intermediate. CP PD PD Parameter B3LYP SCF B3LYP SCF B3LYP SCF r(C1=C2) () 1.295 1.281 1.321 1.320 1.304 1.298 r(C2-C3) () 1.509 1.500 1.466 1.474 1.509 1.510 r(C1-H4) () 1.080 1.070 1.263 1.364 r(C3-H4) () 1.385 1.472 1.093 1.083 (C1C2C3) 64.6 64.7 82.9 85.4 133.1 127.2 The B3LYP/6-311+G(2df,2p) \\ B3LYP/6-31G(d) calculations of this study predicted E 0 =38.48 kcal/mol and H=41.17 kcal/mol for the ring opening step. The reverse of this reaction has an E 0 =15.77 kcal/mol and H=16.87 kcal/mol. Both are in excellent agreement with the previous computational study. This model of the initial ring opening study and Yoshimine’s study are in good agreement with each another and with available experimental evidence. It is known that

PAGE 43

30 DFT often predicts geometries to be slightly larger than experimental, and the transition state geometries are usually surprising late. However this is often considered a price for the highly accurate single point energies that B3LYP/6-311+G(2df,2p) renders. In conclusion this DFT study should be a good model for the study of fluorinated cyclopropenes. The C 3 H 3 F Surface: DFT Study As it has been discussed, fluorine’s influence on the geometry and electronics of cyclopropanes have produced dramatic results. We now wish to extend this knowledge, if only in a theoretical manor, to fluorinated cyclopropenes. The investigation in the rest of this chapter will revolve around the influence of a single atom of fluorine substituted upon the cyclopropene ring. Fluorines substituent effect Through a series of isodesmic equations, Wiberg demonstrated the almost similar stabilization effect gem-difluoro and mono-fluoro substitutions at the methylene carbon have upon cyclopropenes. We extended this work with the following set of isodesmic equations from unweighted B3LYP calculations (Table 2-6). Equation 1 predicts that 1-fluoro substitution gives rise to a significant increase in strain. Where as the 3-fluoro substituent is predicted to be approximately the same as Wiberg’s work.

PAGE 44

31 Table 2-6. Isodesmic of 1-fluorocyclopropene. Eq. Isodesmic H (kcal/mol) 1 H3CFH3CH++FH 8.4 2 H3CFH3CH++FHFF 12.7 3 H2C+HHFH2C+HFH -9.8 Table 2-7 compares the B3LYP\6-31G(d) geometries of cyclopropene, 1-fluorocyclopropene, and 3-fluorocyclopropene. The delocalization of electron density into the bond via the 1-fluoro substituent significantly shortens this bond and the -withdrawing significantly shortens the proximal C-C bond. The result is a lack of electron density in the distal bond, causing a lengthening and likely weakening of this bond. The 3-fluoro substituent causes equal shortening of the framework and some relief of the highly strained angles. Table 2-7. Comparison of fluorinated cyclopropenes. Parameter Cyclopropene 1-Fluorocyclopropene 3-Fluorocyclopropene r(C1-C2) () 1.295 1.287 1.307 r(C1-C3) () 1.509 1.472 1.471 r(C2-C3) () 1.509 1.559 1.471 (C3C1C2) 64.589 68.409 63.624 (C3C2C1) 64.589 61.436 63.624 (H(F)6C3H7) 112.995 113.156 108.169 (C1C3C2) 50.824 50.153 52.753

PAGE 45

32 As discussed in Chapter 1, fluorine to a carbene stabilizes the singlet state compared to the triplet state. -Substitution inhibits 1,2 shifts causing a kinetic stabilization of carbenes. The isodesmic in Figure 2-9 demonstrates the relative effects of versus substitution on the stability of a carbene. H3C H F H H=17.7 kcal/molH3C F H H Figure 2-9. Isodesmic of versus fluorine substitution. An interesting dichotomy has now presented itself. The geometric distortion of 1-fluorocyclopropene suggests the distal bond would be the favored bond broken during homolysis. However, the resulting carbene (substituted) would be much less stable. 1-Fluorocyclopropene The geometry of 1-fluorocyclopropene has already been introduced. The result of fluorine substitution was an increase in strain energy and pointed toward the favorable cleavage of the distal bond. Vinylmethylene Proximal bond breaking leads to two transition states (Figure 2-10). The lower of the two energy barriers (ca. H =29.4 kcal/mol) produces the transoid vinylmethylene with a ca. H=12.3 kcal/mol. The cisoid vinylmethylene has a ca. H =46.5 kcal/mol, and ca. H=16.7 kcal/mol. From the H values, the transoid is the lower energy isomer. Transoid also has the lower barrier to recyclization with a ca. H =17.8 kcal/mol compared to the ca. H =30.5 kcal/mol for the cisoid.

PAGE 46

33 Figure 2-10. Geometries of the proximal vinylmethylene transition states. The geometries in Figure 2-10 also indicate that this transition state occurs earlier in this reaction than that of cyclopropene. This can be seen especially in the small dihedral angle of the methylene hydrogens and the carbon back bone. In the cyclopropene reaction this angle and to a lesser extent the C2-C3 bond length indicates the double bond has significantly formed compared to this case. The vinylmethylene intermediates are depicted in Figure 2-11. The proximal transoid vinylmethylene has C s symmetry. The Proximal cisoid has C 1 symmetry with the C s symmetry intermediate being the transition state connecting the transoid to the cisoid intermediate. Figure 2-11. Vinylmethylene intermediates.

PAGE 47

34 Figure 2-12. Distal bond breaking transition states. Distal bond breaking leads to the transition states found in Figure 2-12. Both the cisoid (DC) and transoid (DT) of C s symmetry are the only stable structures found, and both are transition states. Distal cisoid has a ca. H =29.6 kcal/mol and distal transoid has ca. H =27.0 kcal/mol. The IRC when followed reveals that each of these transition states result in a 180 rotation of the methylene group and reform the cyclopropene. Again these are the only intermediary species found for this distal ring opening and result in an automerization of 1-fluorocyclopropene. Propenylidenes The propenylidene intermediates, although found to have higher energy barriers for ring opening with respect to the vinylmethylenes, were found to be important intermediates for the isomerization of cyclopropene. The synchronous transition states for the propenylidene transition states contains both a 1,2-H shift or 1,2-F shift and cleavage of that bond along the shifting group. Figure 2-13 outlines these transition states.

PAGE 48

35 Figure 2-13. Ring opening schematic of the propenylidene intermediates. For the 1,2-H shifted propenylidene intermediate to cyclize along the same reaction path, a barrier of ca. 9.5 kcal/mol would have to be overcome. However for this intermediate to reach product, the 1-fluoro-2-methylacetylene, a 1,2-F or methyl shift would have to occur. Using the 1,2-F shift leading to the propenylidene intermediate in this system as an example the ca. H =64.7 kcal/mol. Making it an unlikely pathway for 1-fluorocyclopropene to undergo as well as indicating the questionability of it occurring for the aforementioned propenylidene. However, the initial 1,2-F shift does exist on the PES, and if this intermediate were reached it could lead to the 3-fluorocyclopropene system. For the 3-fluorocyclopropene system to be reached the 1,2-F shifted propenylidene intermediate would have to under go a methyl rotation followed by the reverse reaction, but this time with the more favorable hydrogen shift. Cyclization would then lead to the 3-fluorocyclopropene surface.

PAGE 49

36 3-Fluorocyclopropene The structure of 3-fluorocyclopropene was discussed earlier as well as Wiberg’s old and the new DFT isodesmic equations. Propenylidenes As mentioned above one particular pathway from 1-fluorocyclopropene to 3-fluorocyclopropene would be through the propenylidene intermediates. A methyl rotation normally has an energy barrier of about 3 kcal/mol in ethane. With the fluorine substituent effect, this rotation was shown to have a barrier of 4 to 5 kcal/mol on the DFT surface. The 3-fluorocyclopropene path to the propenylidene intermediate is outlined in Figure 2-14. The ring opening has a ca. H =44.8 kcal/mol and the reverse of this reaction has a ca. H =9.5 kcal/mol. Figure 2-14. 3-Fluorocyclopropene to propenylidene schematic. Vinylmethylene The four possible vinylmethylene intermediates are outlined in Figure 2-15. The transoid intermediates have C s symmetry and the cisoids have C 1 . The formation of all intermediates is highly reversible, with the cis-tranoid having the largest ca. H =6.5 kcal/mol. Interconversions of the trans or cis intermediates have ca. H = 16.4 kcal/mol and ca. H =10.0 kcal/mol, respectively. The recyclization barriers are lower than the

PAGE 50

37 interconversion, making the 3-fluorocyclopropene the preferred intermediate among the vinylmethylene intermediates. Figure 2-15. 3-Fluorocyclopropene’s vinylmethylene intermediates. The energy barriers for the ring opening step of 3-fluorocyclopropene are listed in Table 2-8. Cyclopropene’s lowest barrier was found to have a ca. H =36.89 kcal/mol for the ring opening into the vinylmethylene cisoid intermediate. Table 2-8. 3-Fluorocyclopropene ring opening barriers. Intermediate H (kcal/mol) H (kcal/mol) Propenylidene 44.83 35.40 Cis-Cisoid 46.49 43.08 Cis-Transoid 49.49 42.96 Trans-Cisoid 39.56 37.94 Trans-Transoid 37.57 37.33 Conclusions 1-Fluorocyclopropene is predicted to undergo a novel electrocyclic automerization involving a 180 rotation of its methylene group, without the formation of an intermediate

PAGE 51

38 carbene or diradical. The standard deviation from experiments for the B3LYP/6-311+G(2df,2p) // B3LYP/6-31G(d) is documented to be 2.5 kcal/mol. Within this error margin is the calculated energy difference for proximal and distal ring opening mechanisms. Experimental evidence will be difficult to obtain, and any reaction products will probably be obtained from the stabilized proximal ring opening. With the lowest energy barrier for 3-fluorocyclopropene being the ca. H =37.6 kcal/mol for the trans-transoid intermediate, any stabilization imparted by the 3-fluoro substituent does not drastically effect the energy barrier, in absolute terms, for ring opening. The energy barriers for recyclization are also predicted to be slightly higher for the 3-fluorocyclopropene vinylmethylene intermediates compared to the cyclopropene system. Part 2: Difluorocyclopropenes; Computational Study of Structures and Initial Thermal Rearrangement Introduction 3,3-Difluorocyclopropene (Figure 2-16) has been straightforwardly prepared 40 in good yields by dehydrohalogenation of 1-bromo-2,2-difluorocyclopropane. Subsequent vibrational and ab initio analysis has produced the following geometric data seen in Table 2-9. Figure 2-16. The C2v structure of 3,3-difluorocyclopropene.

PAGE 52

39 Table 2-9. Experimental and theoretical geometries of 3,3-difluorocyclopropene. Parameter Exptl 41 B3LYP Calc. r(C1=C2) () 1.321 1.299 r(C1-C3) () 1.438 1.435 r(=C1-H4) () 1.075 1.069 r(C3-F6) () 1.365 1.338 (C1=C2-H5) 148.38 148.54 (F6-C3-F7) 105.48 105.79 (C1-C3(F2)-C2) 54.60 53.83 B3LYP functionals quantitatively predict the following highest occupied (HOMO) and lowest unoccupied (LUMO) molecular orbitals seen in Figure 2-17. -Withdrawing substituents such as fluorine, remove electron density from the bond between C1 and C2 and the antibonding orbital between C3 and C1,C2 as seen in HOMO-2 and LUMO+1. This can be seen in the lengthing of the double bond and shortening of the single bonds (Table 2-10). -Donation would however result in the lengthing of the double bond based upon the LUMO but would have little or no effect upon the single bonds. This analysis based on the MO of cyclopropene has led to the viewpoint that fluorine acts as a -withdrawing substituent in the case of gem-difluorocyclopropene 42 .

PAGE 53

40 Figure 2-17. DFT HOMO’s and LUMO’s of cyclopropene. Table 2-10. Experimental geometries of cyclopropene and 3,3-difluorocyclopropene Parameter Cyclopropene (x=H) Difluorocyclopropene (x=F) r(C=C)() 1.296 1.321 r(C-C) () 1.509 1.438 r(=C-H) () 1.072 1.075 (C=C-H) 149.89 148.38 (X-C-X) 114.66 105.48 (C-C(X2)-C) 50.38 54.60 3,3-Difluorocyclopropene Structure 3,3-Difluorocyclopropene was optimized in the C 2V framework and outlined again in Table 2-11. Again it is still in good agreement with previous literature values.

PAGE 54

41 Table 2-11. B3LYP theoretical geometry of 3,3-difluorocyclopropene. Parameter DFT Calc r(C1=C2) () 1.319 r(C1-C3) () 1.451 r(=C1-H4) () 1.082 r(C3-F6) () 1.366 (C1=C2-H5) 148.0 (F6-C3-F7) 106.2 (C1-C3(F2)-C2) 54.1 Only two vinylmethylene intermediates, cisoid and transoid, are possible upon ring opening. These intermediates and transition state structures are depicted in Figure 2-18. The vinylidene intermediates are C s symmetric and have a very slight preference for the cisoid opening. Figure 2-18. 3,3-Difluoro vinylmethylene intermediate ring opening reactions. The reaction resulting in the propenylidene intermediate is shown in Figure 2-19. With the geminal substituted fluorines on the receiving carbon of the 1,2-H shift, the s-character of the new CF 2 -H bond should be the largest of the propenylidene intermediates. As a result, the C-H bond distance of the transition state was the shortest calculated in the series of cyclopropene, monofluorocyclopropene, and difluorocyclopropene.

PAGE 55

42 Figure 2-19. 3,3-Difluorocyclopropene propenylidene reaction schematic. 1,2-Difluorocyclopropene Interest in the 1,2-difluorocyclopropene system was immediately raised by the results of the 1-fluorocyclopropene rearrangement. However, this time both an and fluorinated carbene would result from the vinylmethylene intermediate ring opening. The 1,2-difluorocyclopropene structure optimized into the Cs point group which resulted in internal angles of 64.9 and 50.2, and the bond lengths shown in Figure 2-20. The C1-C3 bond is the largest reported in the series of cyclopropenes. Figure 2-20. 1,2-Difluorocyclopropene The ring opening mechanism predicted for both cisoid and transoid vinylmethylene intermediates is detailed in Figure 2-21. The transition states are the earliest predicted for the series of cyclopropenes outlined so far. Also, the barriers for cyclization of the vinylmethylene intermediates are also the largest predicted. The energy of reactions coupled with the previous fact indicate these vinylmethylenes would be excellent targets for trapping of the elusive vinylmethylene intermediate

PAGE 56

43 Figure 2-21. 1,2-Difluorocyclopropene vinylmethylene ring opening. The propenylidene intermediate has the most energetic 1,2-F shift of any of the systems. The reaction scheme is seen in Figure 2-22. The high energy barrier and energy of reaction make this propenylidene intermediate an interesting target for trapping. Figure 2-22. The 1,2-difluorocyclopropene to propenylidene rearrangement. Conclusions Wiberg demonstrated that 3,3-difluorocyclopropene was stabilized by almost 10 kcal/mol with respect to cyclopropene. The vinylmethylene ring opening barrier was some 10 kcal/mol higher that cyclopropenes and within the range of barriers seen for 3-fluorocyclopropene. The highly strained 1,2-difluorocyclopropene is predicted to undergo ring opening and form stable vinylmethylene and propenylidene intermediates, the reason for which

PAGE 57

44 must be the -substituted fluorine stabilizing the carbene by increasing the electron density in the C1-C2 bond and the previously demonstrated -fluorination stabilization of carbenes. Although the energy barrier for the ring opening into the propenylidene intermediate is quite large at 65.9 kcal/mol, again the resulting intermediates would make excellent targets for trapping. Part 3: DFT Study of 1-Fluoromethylcyclopropenes Introduction Thermolysis of tetramethylcyclopropene at 260-298C yields two dimethylpenta-1,3-dienes, 2,3-dimethyl-1,3-pentadiene and 2,3-dimethyl -1,3-pentadiene (Figure 2-23). This system leaves little room for interesting single or difluoro substituent effect studies. However, replacing an entire methyl group with fluorine may allow chemists to control the unimolecular pathway. Fundamentally, the more appropriate systems to study theoretically would be 3,3-dimethylcyclopropene of the simpler 3-methylcyclopropene. Experimentally, 1-fluoro-3,3-disubstituted cyclopropenes have been a recent target of synthetic work. 260-298oC + Figure 2-23. Schematic of tetramethylcyclopropene thermolysis. 3-methylcyclopropene and 3,3-dimethylcyclopropene have been shown to rearrange during thermolysis at 185-225C to their respected alkynes, which makes up approximately 90% of the product mixture. The remaining percentage is the respected 1,3-dienes depending upon the starting cyclopropene (Figure 2-24). Both systems

PAGE 58

45 rearrangements are characterized by large negative entropy of activations, presumably corresponding to highly ordered transition states. HHH260-298oCHHH260-298oCH++ Figure 2-24. Schematic of 3-methyl and 3,3-dimethylcyclopropene thermolysis. 1,3,3-trimethylcyclopropene, which when heated to 483 to 525C undergoes a similar ring rearrangement. However, the rate for alkyne formation becomes exceptionally slow and only accounts for 22% of the product mixture (Figure 2-25). H483-525oC++H+ Figure 2-25. Schematic of 1,3,3-trimethylcyclopropene thermolysis. The fall off of alkyne formation can be explained by the favorable formation of the propenylidene intermediate rather than the vinylmethylene intermediate. The propenylidene intermediate then requires a migration of one of the alkyl groups to form the alkyne product. This overall process competes with the vinylmethylene intermediate’s second step of a 1,4-H shift and the formation of the 1,3-diene. The initial vinylmethylene formation, which has been shown to be kinetically favorable, may now lead to the kinetic product. (Figure 2-26).

PAGE 59

46 HHH+ Figure 2-26. Initial ring opening schematic of 1,3,3-trimethylcyclopropene. It has been shown previously that theory predicts the distal bond in 1-fluorocyclopropenes to be the favored bond broken during thermolysis. In the case of any 1-fluoro-methylcyclopropenes the distal bond again is predicted to be the kinetically favored choice for cleavage over the proximal bond. However, in the unlikely case that one of the propenylidene intermediates could compete energetically in the initial ring opening, the following 1,2-F or 1,2-Alkyl shift leading to the alkyne product would still be energetically unfavored. This overall assumption of vinylmethylene preference should also take into consideration that the 3,3-dimethyl system could still allow an initial 1,2-H shift rather than alkyl shift with C2-C3 cleavage. This could prove to be an interesting case study of alkyl versus fluorine migration (Figure 2-27). FRFRRF+RFFRR=Me,H Figure 2-27. 1-Fluoromethylcyclopropenes propenylidene formation schematic.

PAGE 60

47 The more likely energy surface for the 1-fluoromethylcyclopropenes to follow would be for them to undergo the distal bond breaking leading to a vinylmethylene intermediate (Figure 2-28). From there a 1,4-H shift with electronic rearrangement would lead to product formation. In the case of 1-fluoro-3,3-dimethylcyclopropene, the product would be 2-fluoro-1,3-butadiene, and in the case of 1-fluorotrimethycyclopropene, the product would be a similar 1,3-diene. However, depending on the stability of the substituted carbene, the less favorable proximal bond breaking followed by 1,4-H shift could compete over all in the diene formation. FRFR+R=Me,HRFFR+RF Figure 2-28. 1-Fluoromethylcyclopropenes vinylmethylene formation schematic. A theoretical study demonstrating the likelihood that this highly sought after fluorobutadiene could potentially be synthesized in high yields from 1-fluoromethyclopropenes, could be well received by the chemical community. 1-Fluoro-2-methylcyclopropene The optimized structure and selected measurements can be seen in Figure 2-29. Once again an increase of 0.054 is seen in the distal bond, and a smaller 0.039 decrease is seen in the proximal bond. These are slightly larger differences than observed for cyclopropene versus 1-fluorocyclopropene.

PAGE 61

48 Figure 2-29. 1-Fluoro-2-methylcyclopropene optimized structure. The reaction details for the propenylidene intermediates are seen in Figure 2-30. Both intermediates have very large barriers for ring opening via this mechanism, not surprising in the least due to either a 1,2-F or 1,2-Methyl shift. However, this is the first time a proximal bond has been found to be the favored bond broken during a ring opening step. Figure 2-30. 1-Fluoro-2-methylcyclopropene to propenylidene intermediate schematic. The vinylmethylene intermediates are outlined in Figure 2-31. The distal ring opening leading to the transoid form has the lowest ca. H = 25.4 kcal/mol. The distal cisoid and proximal transoid cleavages are competitive with ca. H values of 28.0 and 28.4 kcal/mol, respectfully, followed by a ca. H = 45.0 kcal/mol for the proximal cisoid ring opening.

PAGE 62

49 Figure 2-31. 1-Fluoro-2-methylcyclopropene, Vinylmethylene intermediates. Distal ring openings have barriers of less than 4 kcal/mol for ring closure and proximal transoid has a barrier of ca. H = 13.9 kcal/mol. By far the proximal cisoid cyclization ca. H =26.5 makes this the most stable intermediate of the vinylmethylenes and has the second smallest ca. H= 19.57 kcal/mol. The smallest ca. H=15.39 kcal/mol is for the proximal transoid intermediate. 1-Fluoro-3,3-dimethylcyclopropene 1-Fluoro-3,3-dimethylcyclopropene optimized in the C s point group and has a structure very similar to 1-fluoro-2-methylcyclopropene (Figure 2-32). The energy barrier for the rotation of a single methyl group in this cyclopropene system was found to be 1.53 kcal/mol.

PAGE 63

50 Figure 2-32. 1-Fluoro-3,3-dimethylcyclopropene. The propylidene intermediates where found to lie 22.86 kcal/mol and 36.39 kcal/mol above 1-fluoro-3,3-dimethylcyclopropene. The lower lying intermediate was the structure arising from a fluorine migration. However a direct transition state for the fluorine transfer during bond homolysis was not found. A transition state with a ca. H =43.44 kcal/mol for the hydrogen transfer was found. Of the four vinylmethylene intermediates, the two arising from the distal bond breaking are favored by approximately 5 kcal/mol. A ca. H =20.95 kcal/mol is the lowest leading to the distal cisoid structure, with a slightly higher ca. H =21.06 kcal/mol for the transoid structure. This transoid intermediate also has the higher energy barrier for recyclization with a ca. H = 19.77 compared to cisoids ca. H= 12.70 kcal/mol. The proximal intermediates have energy barriers that were isoenergetic with a ca. H =26.74 kcal/mol. The proximal cisoid intermediate was some 7 kcal/mol lower in energy than the transoid with a ca. H= 12.70 kcal/mol. In both the distal and proximal

PAGE 64

51 ring openings, the cisoid intermediates were the lower energy intermediates which is unfortunate in that product formation arises from the transoid form. However, the cisoid intermediates also have the lowest barrier to recyclize, with the distal cisoid having a barrierless route, and the proximal transoid has a ca. H =14.8 kcal/mol. Direct transition states for the conversion of cisoid to transoid intermediates were calculated for both the proximal and distal ring openings. This was done to determine whether this was the more likely route versus the cyclopropene opening directly to the transoid intermediate. Distal cisoid to distal transoid has a ca. H = 24.3 kcal/mol where the reverse has a barrier of 27.0 kcal/mol. For the proximal cisoid to transoid interconversion the ca. H = 6.5 kcal/mol and the reverse is 13.5 kcal/mol. Once the transoid intermediates are obtained in either case, a 1,4 H-shift leads to the 1,3-butadiene products. In the case of proximal ring opening, 1-fluoro-3-methyl-1,3-butadiene would be the product. The more favored 2-fluoro-3-methyl-1,3-butadiene arises from the distal ring opening. As for the distal transoid to 2-fluoro-3-methylbutadiene transition state, a single negative frequency of -1503.1 cm -1 was found for the transition state with a barrier of ca. H = 14.29 kcal/mol leading to product. In the case of proximal transoid to 1-fluoro-3-methylbutadiene transition state, a single negative frequency of -1408.4 -1 was found for the transition state with a ca. H = 16.34 kcal/mol leading to product. Conclusions For the first time ring opening of the proximal bond is seen as the favorable route for the 1-fluoro-2-methylcyclopropene ring opening into the propenylidene intermediate. However, this route is considerably higher in energy compared to the vinylmethylene

PAGE 65

52 intermediate. The kinetic and thermodynamic vinylmethylene intermediates, the cisoid forms, are dead ends with regard to butadiene formation. The transoid forms must be obtained for the formation of the 1,3-butadiene products. The large negative frequency for the transition states of the 1,4-H shift indicates some tunneling is occurring to reach products.

PAGE 66

CHAPTER 3 2-FLUOROCYCLOPROPENYLCARBINYL RADICAL AND CATION REARRANGEMENTS Introduction The Dolbier group recently completed the experimental and theoretical 12, 43-45 examination of the 2,2-difluorocyclopropylcarbinyl radical and cation ring opening (Figure 3-1). Results of this study lead to the documentation of rates, energy barriers, and demonstration of this as a new hypersensitive probe. CCC C F F H H H H H CC C F F H H H H H CC C H H H H H F F CCC C F F H H H H H Figure 3-1. 2,2-Difluorocyclopropylcarbinyl ring opening. It was previously predicted 12, 43-45 that the gem-difluorocyclopropylcarbinyl cation was not a minimum on several PES and contains no activation barrier for the rearrangement to the 1,1-difluoro-homoallylic cation. Solvolytic studies on model systems also confirmed this prediction. In contrast, the 2,2-difluorocyclopropylcarbinyl radical was found to be a stable intermediate on several computational levels of theory. The DFT study predicted an energy barrier of 2.2 kcal/mol for the distal bond breaking leading to the 2,2-difluoro-3-buten-1-yl radical. The barrier is approximately twice as large for the 2-fluorocyclo-propylcarbinyl radical and decreases by some 75% when a 53

PAGE 67

54 methyl substituent is added to the carbonyl position. Computationally, the rate for ring opening was predicted to be 1.5x10 11 s -1 at 25C and was experimentally determined at 99.3C to be 1.3x10 11 s -1 . Cation The C 4 H 5 + surface has been the subject of many computational studies 46-49 due to the aromatic ion (1), homoaromaticity studies, and general classical/nonclassical cation structure debate. Recently, the cyclopropenylcarbinyl cation (Figure 3-2) was studied at the HF and MP2 level. 50 It was predicted that the cyclopropenylcarbinyl cation geometry is a transition state (TS1) for the methylcyclopropenyl cation (1) to classical vinyl cation (2) rearrangement. Ultimately intermediate (2) leads to the cyclobutene-3-yl cation (3) via a transition state (TS2), which is best described as a bicyclobutenium ion. CH3 H H H H H H H H H H H H H H H H H H H H H H (1)(TS1)(2)(TS2)(3)1.647 Figure 3-2. Cyclopropenylcarbinyl schematic. Another interesting intermediate to note on this PES (Figure 3-3) is the cyclobutenyl cation (5) resulting from a two step process. First, a transition state (TS3)

PAGE 68

55 approximately 20 kcal/mol higher than the aforementioned (TS2), forms a classical vinyl cation (4), which is stabilized by -donation from the vinyl substituent. Second, Another 23.3 kcal/mol energy barrier must then be crossed which leads to (5). This final transition state (TS4) has the geometry consistent with a nonclassical carbonium ion structure. The high energy barriers for the six lowest lying cations make them good candidates for observation in the gas phase. Figure 3-3. PES schematic of the cyclopropenylcarbinyl rearrangement 50 . Radical Ab initio studies 51, 52 confirmed the global minima of the C 4 H 5 radical surface to be the H 3 C 4 H 2 conformer (figure 3-4). None of these studies mention the cyclopropenyl-carbinyl radical. As a matter of fact, this strained isomer and the fused bicyclopropyl,

PAGE 69

56 have only been identified in solid matrices via ESR spectroscopy. 53 Thermodynamic data is missing from the ESR study and all other literature concerning these two isomers. H3C H3C Figure 3-4. Global minima of the C 4 H 5 radical. Computational Detail All structures were optimized using density functional theory (DFT) calculations within the Gaussian 98 and 03 program packages. The level of theory was set to the unrestricted Becke’s hybrid three-parameter functional (UB3LYP) for open shell (radical) species and B3LYP for closed shell species, and in both cases the 6-31G(d) basis set was used. Frequency calculations were performed on all stationary points to identify transition structures and to determine thermochemical information. Transition structures were characterized by a single imaginary frequency. Thermal energies and entropy terms were obtained using frequencies scaled by 0.9807 at 298.15 K and 1 atm. An intrinsic reaction coordinate (IRC) calculation was performed for each transition structure to follow the entire reaction pathway for each transition structure. Single-point energies were calculated using uB3LYP level of theory and the 6-311+G(2df,2p) basis set. Calculated rate constants were derived from transition state theory. The Cyclopropenylcarbinyl Cation: A DFT Study Starting with a guess for the geometry of the primary cation, (cyclopropenyl-3-carbinyl), an exhaustive search resulted in the location of a transition state with a

PAGE 70

57 negative frequency of -788.33cm -1 . This transition state when followed to products resulted in the aromatic methylcyclopropenyl cation [1] global minima, and the 1,3-butadien-1-yl [2] intermediate. These structures are depicted with bond length detail in Figure 3-5. Figure 3-5. Cyclopropenylcarbinyl Structures. The carbon backbone dihedral angle and the normal bond angles are listed in Table 3-1. The reaction starts with a bending of the carbon backbone and when a methyl hydrogen becomes perpendicular to the C1-C4 bond, it can shift allowing the positive charge to accumulate in the primary position. This flat transition state geometry relaxes into the vinyl cation by allowing the system to delocalize. Table 3-1. Angles and dihedrals of the cyclopropenylcarbinyl ring opening. Structure (C1C2C3) (C2C3C1)(C2C1C3)(C4C1C2) (C3C2C1C4) [1] 60.5 60.5 59.0 150.6 180.0 [TS1] 63.0 65.3 51.7 149.9 140.3 [2] 77.0 58.9 44.1 122.0 121.0 This first step (Figure 3-6), going from the aromatic system [1] to what was the vinyl cation (2) of the HF/MP2 study. In this B3LYP surface the transion state [TS1] leads to what would be better described as a 1,3-butadien-1-yl cation, with a ca. H =60.9 kcal/mol. The reverse, sequential cyclization and hydrogren shift has a ca. H =25.5 kcal/mol. The following transition state [TS2] however has ca. H =0.3 kcal/mol barrier to cyclize to the cyclobutane-3-yl cation [3]. The energy barrier is

PAGE 71

58 considerably lower than that of the previous study, and so low that the transition state structure [TS2] is obviously very similar to [2] and the resulting cyclobutenyl cation [3] is seen in Figure 3-7. With the denoted distance of 1.772 across the ring, this cation owes its stability to homoaromaticity. Qualitatively this can been seen in Figure 3-8 indicated by the large p-orbital of the HOMO from the DFT calculation. Figure 3-6. The B3LYP PES schematic of the cyclopropenylcarbinyl rearrangement Figure 3-7. Cyclobutenyl cation structure.

PAGE 72

59 Figure 3-8. HOMO of the 2-cyclobutenyl cation. 2-Fluorocyclopropenylcarbinyl Cation: A DFT Study Gas Phase ually replacing the hydrogens at the C2 and C3 carbons of the primary cationo e Figure 3-9. The two initial 2-fluorocyclopropenylcarbinyl cation transition states Individ transition state with a fluorine started the examination of the C4H4F + PES. Twtransition states were found for the respective distal and proximal bond opening. Both ardepicted in Figure 3-9. Upon inspection of the methylene carbon, one would expect the empty p-orbital to be aligned with the proximal bond in the case of [1] and to be alignedwith the distal bond in [2]. The transition state geometries are rather flat and from the position of the hydrogen on C3 the geometry contains the start of the hydrogen shift.

PAGE 73

60 Upon examination of the IRC’s for both transition states, [1] resulted in the formation of 1-fluorocyclobutenyl and [2] resulted in 2-fluoro-cyclobutenyl cation (Figure 3-10). Both led to the global minimum of the aromatic 1-fluoro-3-methylcyclopropenyl cation (Figure 3-11). The 1-fluorocyclobutenyl cation is a planar cation, which has the positive charge stabilized on the carbon bearing the fluorine. This type of stabilization was discussed in chapter 1. The 2-fluorocyclobutenyl cation benefits from the homoaromaticity as in the Figure 3-10. Bond distances of 1-fluoro and 2-fluorocyclobutenyl cations Figure 3-11. The 1-Fluoro-3-methylcyclopropenylcarbinyl minima

PAGE 74

61 hydrocarbon examthe fluorine and carbon ol above the aromol higher than the 1-fluoro cation, and 12 kcal/mtops in that goes eometry and no transition states were found in those regions. ple. While the hydrocarbon [3], had a dihedral angle of the butenyl ring of 26.5, the fluorocarbon has a greater bend with a dihedral of 35.9. The HOMO is depicted in Figure 3-12. As one can see, a node exists between ring, restricting the space for delocalization. Of the two fluorocarbons, the 1-fluorocyclobutenyl cation was found to be the lower energy cation, lying 2 kcal/matic ion. The 2-fluorocyclobutenyl cation was 10 kcal/mol higher than the aromatic ion. Figure 3-12. The HOMO of 2-fluorocyclobutenyl cation. The PES (Figure 3-13) from the IRC calculations shows two areas of leveling off on either side of the transition states. In the case of [1], the IRC calculation sregion due to the broad flat surface. On the other side, with [2], the IRC calculation completely to the 2-fluorocyclobutenyl cation. Examination of these flat surfaces revealed unstable intermediates of the vinyl cation/1,3-butenyl g

PAGE 75

62 h ocyclobutenyl cation is the favored cycloprope cyclopropcalcukcal/m Figure 3-13. The 6-31G(d) PES of the IRC calculations Transition state [1], which corresponds to a distal ring opening/expansion, was found to have a ca. H =62.3 kcal/mol. This was somewhat lower than the ca. H =64.6 kcal for transition state 2, a proximal ring opening/expansion. So, botkinetically and thermodynamically the 1-fluor pathway, which again derives from chemistry of the distal bond. Transition states [1] and [2] in essence have the p-orbital eclipsing the ene ring. The search for a transition state having the p-orbital bisecting thene ring resulted in transion state [3] depicted in figure 3-14. The IRC lations showed the resulting products of [3] are the 1-fluoro (ca. H=37.7 ol) and 2-fluorocyclobutenyl cations (ca. H=28.2 kcal/mol).

PAGE 76

63 Figure 3-14. The bisected transition state. Solvent Phase Transition states [1], [2], and [3] were now optimized using the Polarizable Model (PCM) with the solvent set as water. Of the three transition state etry inputs, only two transition states were found. These two are depicted in figure 3-15. The bisected structure of the gas phase and solvent phase are of considerably ilar geometries. As before the bisected transition state connects the 1-fluoro and 2-fluorocyclobutenyl cations, as does the new eclipsed transition state. The eclipsed Continuumgeomsimol higher in energy and the comparison of the in Figure 3-16. transition state is approximately 15 kcal/m IRC PES is seen

PAGE 77

64 Figure 3-15. The eclipsed and bisected transition states of the PCM calculations. Figure 3-16. PCM-IRC of the bisected and eclipsed transition states. Cyclopropenylcarbinyl Radical: A DFT Study Starting with the cyclopropenylcarbinyl radical the barrier for ring opening (Figure ) was determined to be ca. H=5.2 kcal/mol. This is 1.2 kcal/mol lower than the cyclopropylcarbinyl radical ring opening44at the same level of theory. The resultin 3-17, Path Ag

PAGE 78

65 ca. H=-17.35 kcal/mol is still some 9.5 kcal/mol above the cyclobute nyl radical. A barrier of 36.9 kcal/um. tion kcal/mol. The final step is a ring openi mol would have to be overcome to reach that minim Figure 3-17. Cyclopropenyl Radical schematic. Another route available to the cyclopropenylcarbinyl radical {1} is the cyclizainto the exo-bicyclobutanyl radical (Figure 3-17, Path B). Although this ca. H=21.0 ol is not competitive with the ring opening, the resulting bicyclic product is of theoretical interest and hence further explored. The exo-bicyclobutanyl radical can ring expand with a ca. H =8.2 kcal/mol and a ca. H=-26.6 kcal/m ng into the 1,3-butadienyl radical, which would lead to the global minimum on this radical surface.

PAGE 79

66 2-Fluorocyclopropenylcarbinyl Radical: A DFT Study Distal Ring Opening The 2-fluorocyclopropenylcarbinyl radical can undergo ring opening via either distal or proximal bond breaki ng. In both cases the resultant bicyclobutanyl radicals were explored. In Figrrows indicate the reverse reactions. radical {. This is an order of mon of the exn ure 3-18, which outlines the distal ring opening, the inner a the energy barriers for Figure 3-18. 2-Fluorocyclopropenylcarbinyl radical, distal bond rearrangements. Again the favored path is the ring opening into the 2-fluoro-1,3-butadien-1-yl 2}. The ca. H=0.40 kcal/mol results in a reaction rate of 7.68x1012s-1agnitude faster than the difluorocyclopropylcarbinyl radical. The other option for {1} is significantly higher with a ca. H =19.8 kcal/mol for the formati o-bicyclic radical {3}. IRC calculations indicated that in actuality this transitiostate leads to the endo-bicyclic radical (figure 3-19). The endo to exo transition state

PAGE 80

67 (figure 3-20) has a ca. H =13.0 kcal/mol, some 6 kcal/mol lower than the transition for {1} leading to the endo-bicyclic radical. The exo-isomer (figure 3-21) is ca. 5.05 kcal/mol lower in energy. The transition state for the endo to exo involves the flipping the fluorine atom which occurs close to planarity of the fluorine atom. The overall conclusion drawn from these points on the PES is that once the endo-bicyclic ring is formed, it immediately rearranges to the exo-isomer. state of Figure 3-20. The endo-exo transition state. Figure 3-19. The endo-1-fluoro-bicyclobutanyl radical.

PAGE 81

68 yl radical {3}. kr=9.43x10s, some five orders of magnitude lower than that of the distal. Figure 3-21. The exo-1-fluoro-bicyclobutan The bicyclobutanyl radical {3} would likely expand into the 2-fluorocyclobutenyl radical {4} with a ca. H=9.46 kcal/mol and a H=-29.8 kcal/mol. This cyclobutenyl radical {4} again sits at the minimum on this limited surface, but a ca. H=38.02 kcal/mol leads to the ring opening, 2-fluoro-1,3-butadien-1-yl radical, which ultimately would lead to the global minima. Proximal Ring Opening Figure 3-22 outlines the proximal ring opening surface. Again, the direct opening into the 1-fluoro-1,3-butadienyl radical {5} is favored, but the energy barrier is almost 7 kcal/mol above that for the distal ring opening. The rate for this ring opening is ca. 7-1

PAGE 82

69 Figure 3-22. The proximal ring opening schematic. Further exploration of the reactions centered around the proximal bond reveal a transition state with a ca. H=24.1 kcal/mol, leading to the endo isomer. The energy point calculation of this isomer revels a geometry (figure 3-23) some 61 kcal/mol higher in energy than the 2-fluorocyclopropenylcarbinyl radical. This study finds the exo isom er ergy compared to the endo, and 2.5 kcal/mer {6} (figure 3-24) to be some 64 kcal/mol lower in en ol lower in energy than the cyclopropenylcarbinyl radical. From this limited exploration of the PES, it is believed that the transition state that leads to the endo-isomactually leads directly to the exo isomer.

PAGE 83

70 Figure 3-23. The endo-2-fluoro-bicyclobutenyl radical. Figure 3-24. The exo-2-fluoro-bicyclobutenyl radical {6}. A ca. H =6.99 kcal/mol barrier leads to the ring expanded, 1-fluorocyclobutenyl radical {7}. Now, two reaction paths present themselves for ring opening. The first occurs proximal to the fluorine bond, with a ca. H =33.89 kcal/mol, resulting in {5}. The other has a ca. H =35.0 kcal/mol and leads to the lower energy 3-fluoro-1,3-butadien-1-yl radical {8}. Both of these energy barriers and the ca. H =36.7 kcal/mol

PAGE 84

71 going back to the bicyclobutanyl radical are large enough that {7} might make for a good trapping target. An attempt to explain the unusually high energy of the endo-2-fluorobicyclo-butanyl radical leads to the examination of the SOMOs of the hydrocarbon (figure 3-25), 2-fluoro-bicyclobutan-1-yl (figure 3-26), and 1-fluoro-bicyclobutan-1-yl (figure 3-27) radicals. The figures below are rotated around the z-axis, which was considered to run through the middle of the carbon ring. Figure 3-25. SOMO of the bicyclobutenyl radical Figure 3-26. SOMO of the 2-fluoro-bicyclobutenyl radical

PAGE 85

72 Figure 3-26. Continued Figure 3-27.SOMO of the 1-fluoro-bicyclobutenyl radical In the hydrocarbon and 1-fluorobicyclic structures, the endo and exo-isomers allowed the orbital centred on the carbon, containing the -radical, to interact with the bisected edge of the cyclopropanes, thus providing delocalization into the whole system. In the case of the 2-fluoro-bicyclic structures the p-orbital of the fluorine is interacting with the edge on orbitals, forcing the radical to become totally -type and localized on the carbon. Although the radical does not account for the additional energy, the donation of the fluorine into the strained bicyclic structure may, and further examination would be beneficial to the understanding and explanation of the destabilized endo-1-fluorobicyclobutan-1-yl radical.

PAGE 86

73 Conclusions The theoretical explorations of the hydrocarbon cyclopropropenylcarbinyl cation and radical rearrangements were compared to those of the 1-fluorocyclopropenyl-3-carbinyl cation and radical. Both cases showed the remarkable influence fluorine can have on the geometry and reaction dynamics of these systems. In the cation example, it was shown that the global minima in all cases are the aromatic cyclopropenyl cations. However, depending upon which side of the PES one uses to enter the 2-fluorocyclopropenylcarbinal cation surface, the planar 1-fluorocyclobutenyl cation or the homoaromatic 2-fluorocyclobutenyl cation would be formed. In either case, products would arise from the trapping of the cyclobutenyl cations. The radical system was shown to rearrange with preference towards ring opening and forming the 1,3-butadien-1-yl radical. The 1-fluorocyclopropenyl-3-carbinyl radical exploration resulted in finding the very fast and favorable opening of the distal bond to fluorine to form the products of this surface. These would arise from the 2-fluoro-1,3-butadien-1-yl radical. Further exploration of the bicyclic and cyclobutenyl radicals of this surface could result in some interesting targets for trapping.

PAGE 87

74 REFERENCES 1. CRC Handbook of Chemistry and Physics. 84th ed.; CRC Press: Boca Raton, 2003-2004. 2. Molecular structure and energetics. VCH Publishers: Deerfield Beach, FL, 1986. 3. Hudlicky, M.; Pavlath, A. E., Chemistry of organic fluorine compounds II : a critical review. ed.; American Chemical Society: Washington, DC, 1995. 4. Harmony, M. D.; Laurie, V. W.; Kuczkowski, R. L.; Schwendeman, R. H.; Ramsay, D. A.; Lovas, F. J.; Lafferty, W. J.; Maki, A. G., Journal of Physical and Chemical Reference Data 1979, 8, (3), 619-721. 5. Typke, V.; Dakkouri, M.; Oberhammer, H., Journal of Molecular Structure: THEOCHEM 1978, 44, (1), 85-96. 6. Pople, J. A.; Radom, L.; Hehre, W. J., The Journal of The American Chemical Society 1971, 93, (2), 289-566. 7. Bastiansen, O.; Hassel, O.; Tidskkr; Kjemi; Bergues, Metallurgi 1946, 6, 71. 8. Rappoport, Z., The Chemistry of the cyclopropyl group.; Wiley: Chichester West Sussex ; New York, 1995. 9. Yamamoto, S.; Nakata, M.; Fukuyama, T.; Kuchitsu, K., The Journal of Physical Chemistry 1985, 89, (15), 3298 3302. 10. Wendoloski, K. B. W. a. J. J., The Journal of The American Chemical Society 1982, 104, (21), 5679 5686. 11. Kresge, C. H. A. a. A. J., The Journal of The American Chemical Society 1986, 108, (25), 7912-7918. 12. Tian, F. Experimental and computational investigations of the effects of gem-difluoro-substituents on cyclopropane. 1999. 13. Egawa, T.; Fukuyama, T.; Yamamoto, S.; Takabayashi, F.; Kambara, H.; Ueda, T.; Kuchitsu, K., The Journal of Chemical Physics 1987, 86, (11), 6018-6026.

PAGE 88

75 14. Greenberg, A.; Liebman, J. F., Strained organic molecules. ed.; Academic Press: New York, 1978. 15. Walsh, A. D., Transactions of the Faraday Society 1949, 45, (2), 179-190. 16. Forster, T., Zeitschrift Fur Naturforschung Section B-a Journal of Chemical Sciences 1939, 43, 58. 17. Coulson, C. A.; Moffitt, W. E., Philosophical Magazine 1949, 40, (300), 1-35. 18. West, S. W. T. a. R., The Journal of The American Chemical Society 1966, 88, (11), 2481-2488. 19. Stigliani, W. M.; Laurie, V. W.; Li, J. C., The Journal of Chemical Physics 1975, 62, (5), 1890-1892. 20. Tidwell, T. T., Angewandte Chemie-International Edition 1984, 23, (1), 20-32. 21. Walsh, I. M. B. a. R., Journal of the Chemical Society-Faraday Transactions I 1978, 74, 1146. 22. Walsh, R., Journal of the Chemical Society-Faraday Transactions I 1976, 74, 2137-2138. 23. Hopf, H.; Priebe, H.; Walsh, R., The Journal of The American Chemical Society 1980, 102, (3), 1210-1212. 24. Kiefer, J. H.; Kumaran, S. S.; Mudipalli, P. S., Chemical Physics Letters 1994, 224, (1-2), 51-55. 25. Karni, M.; Oref, I.; Barzilai-Gilboa, S.; Lifshitz, A., The Journal of Physical Chemistry 1988, 92, 6924-6929. 26. Srinivasan, R., The Journal of The American Chemical Society 1969, 91, (23), 6250-6252. 27. Srinivasan, R., Journal of the Chemical Society D-Chemical Communications 1971, (17), 1041-1042. 28. Hopf, H.; Wachholz, G.; Walsh, R., Chemische Berichte-Recueil 1985, 118, (9), 3579-3587. 29. Hopf, H.; Wachholz, G.; Walsh, R., Journal of the Chemical Society-Perkin Transactions 2 1986, (7), 1103-1106.

PAGE 89

76 30. York, E. J.; Dittmar, W.; Stevenso.Jr; Bergman, R. G., The Journal of The American Chemical Society 1973, 95, (17), 5680-5687. 31. Honjou, N.; Pacansky, J.; Yoshimine, M., The Journal of The American Chemical Society 1989, 111, Ab Initio Studies of the C 3 H 4 Surface. 32. Minato, T.; Osamura, Y.; Yamabe, S.; Fukui, K., The Journal of The American Chemical Society 1980, 102, (2), 581-589. 33. Pasto, D. J.; Haley, M.; Chipman, D. M., The Journal of The American Chemical Society 1978, 100, (17), 5272-5278. 34. Hutton, R. S. M., M.L.; Roth, H.D.; Wessermann, E., The Journal of The American Chemical Society 1974, 96, (14), 4680. 35. Costain, C. C., The Journal of Chemical Physics 1958, 29, (4), 864 888. 36. Toth, A. G. M. a. R. A., Journal of Molecular Spectroscopy 1965, 17, 136-155. 37. Honjou, N.; Pacansky, J.; Yoshimine, M., The Journal of The American Chemical Society 1989, 111, 2785-2798. 38. Honjou, N.; Pacansky, J.; Yoshimine, M., The Journal of The American Chemical Society 1985, 107, 5332-5341. 39. Honjou, N.; Pacansky, J.; Yoshimine, M., The Journal of The American Chemical Society 1984, 106, 5361-5363. 40. Craig, N. C.; MacPhail, R. A.; Spiegel, D. A., The Journal of Physical Chemistry 1978, 82, (9), 1056-1070. 41. Ramaprasad, K. R.; Laurie, V. W., The Journal of Chemical Physics 1976, 64, (12), 4832-4835. 42. Lien, M. N.; Hopkinson, A. C., Journal of Molecular Structure: THEOCHEM 1987, 149, 139-151. 43. Tian, F.; Battiste, M. A.; Dolbier Jr., W. R., Organic Letters 1999, 1, (2), 193-195. 44. Tian, F.; Dolbier Jr., W. R., Organic Letters 2000, 2, (6), 835-837. 45. Tian, F.; Lewis, S. B.; Bartberger, M. D.; Dolbier, W. R.; Borden, W. T., The Journal of The American Chemical Society 1998, 120, (24), 6187-6188.

PAGE 90

77 46. Dorado, M.; Mo, O.; Yanez, M., The Journal of The American Chemical Society 1980, 102, (3), 947-950. 47. Apeloig, Y.; Collins, J. B.; Cremer, D.; Bally, T.; Haselbach, E.; Pople, J. A.; Chandrasekhar, J.; Schleyer, P. V. R., The Journal of Organic Chemistry 1980, 45, (17), 3496-3501. 48. Mayr, H.; Schneider, R.; Wilhelm, D.; Schleyer, P. V., The Journal of Organic Chemistry 1981, 46, (26), 5336 5340. 49. Lien, A. C. H. a. M. H., The Journal of The American Chemical Society 1986, 108, (1), 2843-2849. 50. Cunje, A.; Rodriquez, C. F.; Lien, M. H.; Hopkinson, A. C., The Journal of Organic Chemistry 1996, 61, 5212-5220. 51. Cooksy, C. L. P. a. A. L., The Journal of Physical Chemistry A 1998, 102, (30), 6186-6190. 52. Cooksy, C. L. P. a. A. L., The Journal of Physical Chemistry A 1999, 103, 2160-2169. 53. Kaiser, R. I.; Stranges, D.; Bevsek, H. M.; Lee, Y. T.; Suits, A. G., The Journal of Chemical Physics 1997, 106, (12), 4945-4953.

PAGE 91

BIOGRAPHICAL SKETCH Born February 11, 1974, Bob spent most of his formidable years outside a small rural Illinois town within the shadow of the St. Louis Arch. Upon graduating Freeburg Community High School, home of the Mighty Midgets, in 1992 Bob enrolled at Southern Illinois University – Edwardsville (SIUE). He spent the first 3 years at SIUE studying genetics before switching to chemistry. Upon graduation in 1997, after spending a year doing undergraduate research in biochemistry, Bob continued as a Cougar at SIUE under the tutelage of Dr. Tim Patrick, improving upon the synthesis of a fluorinated synthon, and showing its use in the synthesis of (Z)-5-(1-Decenyl)dihydro-2(3H)-furanon, a potential pheromone for the Japanese Beetle. Continuing his educational endeavors, Bob enrolled at the University of Florida to work with Dr. William R. Dolbier, a world leader in physical fluoro-organic studies. While Bob’s natural interests have always led him to explore the potential use of computers in many scientific applications, it wasn’t until being introduced to computational chemistry in an investigation of the unimolecular rearrangement of 1-fluorocyclopropene that Bob realized his potential in the field. Bob also spent his years as a Gator working as a research assistant for Dr. Khalil Abboud at the Center for X-ray Crystallography. Bob designed and implemented a data management and distribution system while learning the basics of crystallography. 78

PAGE 92

79 Upon graduation, Bob plans to pursue post doctoral study with Weston Borden, the new Welch Chair in Chemistry, at the University of North Texas, home of the NCAA Divison 1-A, Mean Green. And that is how one becomes a Mean Green Gator Midget.