Ultrafast Time Resolved and Computational Studies of and Diazirine Excited

States, and of

DISSERTATION

Presented in Partial Fulfillment of the Requirements for

the Degree Doctor of Philosophy in the Graduate

School of The Ohio State University

By

Yunlong Zhang, M.S.

Graduate Program in Chemistry

The Ohio State University 2010

Dissertation Committee:

Professor Matthew S. Platz, Advisor

Professor Christopher M. Hadad

Professor T. V. RajanBabu

Copyright by

Yunlong Zhang

2010

ABSTRACT

Ab initio quantum calculations and ultrafast time-resolved laser flash photolysis

techniques have been used to study singlet carbenes and the photochemistry of diazirines

and diazo compounds. After a brief introduction of chemistry in Chapter 1, the

photophysics and photochemistry of aryldiazirines are investigated in Chapters 2 through

6. Detailed theoretical calculations begin with parent phenyldiazirine and its isomer

phenyldiazomethane. The structures of the ground and electronic excited states (S 1, S 2,

and S 3) of both compounds are optimized with RI-CC2 and DFT methods. The denitrogenation of both phenyldiazirine and phenyldiazomethane to produce singlet phenylcarbene, and the isomerization between both compounds, are investigated mechanistically on their potential energy surfaces. These predictions support the spectroscopic assignment in ultrafast studies of arylalkyldiazirines in chapters 3 – 6 and the accuracy of these theoretical methods are calibrated by the excellent agreement with experimental data. In Chapter 3 we present the first direct observation of singlet phenylcarbene and measurement of its lifetime in solution using ultrafast time-resolved infrared spectroscopy. In Chapter 4 we provide the first direct observation of the S 1 excited state of para -methoxy-3-phenyl-3-methyl diazirine ( p-CH 3OC 6H4CN 2CH 3) with

both IR and UV–vis detection techniques. The S1 state of the diazirine decays into the

ii diazo compound directly. The S 2 excited state is populated with 270 nm light and decays

directly into singlet arylcarbene and diazo compound, as well as the S 1 state, via internal

conversion. A Hammett study of the S 1 excited states is discussed in Chapter 5. An excellent linear correlation is obtained between the S 1 lifetimes of arylchlorodiazirines

+ and their para - substituent σp parameters. The effect of substitution of β- on the S 1 state lifetimes is examined in Chapter 6 and is consistent with the RIES

mechanism. The wavelength dependence effect on the photochemistry of aryldiazirines

was discussed.

In Chapter 7 we present the first direct observation of a singlet vinylcarbene and

study its cyclization to a product in solution. Calculations predict that

singlet vinylcarbene is highly delocalized over the C=C double bond.

The photochemistry of N,N-diethyldiazoacetamide is detailed in Chapter 8. We

concluded that the excited state of the diazoamide precursor undergoes direct

intramolecular C−H insertions in forming both β- and γ -lactams, as well as denitrogenation to produce singlet carbene. The relaxed singlet carbene decays by isomerizing into γ-lactam in chloroform, and in , this path is suppressed by intermolecular OH insertion reactions.

iii

DEDICATION

Dedicated to my parents and my wife Na

iv ACKNOWLEDGMENTS

I would like to express my sincerest gratitude to my advisor, Dr. Matthew S.

Platz. I am very lucky to be his student. His dedication, vision, and enthusiasm in

research inspired me during my study at The Ohio State University. His guidance and

encouragement has been instrumental in my studies, and his mentoring of both my

research and personal development has been invaluable.

I would like to thank Dr. Christopher M. Hadad and Dr. T.V. RajanBabu for serving on my dissertation committee and their insights and suggestions and endless help with my research through the years.

I also enjoy the collaboration with Dr. Robert A. Moss at Rutgers and Dr. Guy

Buntinx at Lille and thank them for their stimulating discussions.

It has been a great honor and pleasure to work with the members in the Platz group. Special thanks go to the two gifted physicists Dr. Jacek Kubicki and Dr. Gotard

Burdzinski. Both have given me tremendous assistance. I am also grateful to Dr. Jin

Wang who has offered me endless advice and help on my initial days in the lab. I also thank Dr. Arthur Winter and Mr. Shubham Vyas for their help on the calculations.

Furthermore, I would like to thank the help and friendship from in and out of the Platz

v group, particularly Drs. Peter Selvaraj, Biswajit Saha, Huolei Peng, Jiadan Xue, Ozlem

Dogan Ekici, Chris Middleton, Muthukrishnan Siva, Michel.Sliwa, and Mr. Calvin Luk

Beyond my fellow co-workers, I would like to acknowledge my best friends in

Columbus, Yujie Sun, Wenlan Chen, Siyu Tu, Xiaozhao Wang and Chuang Tan. I am extremely grateful for their friendship, help, encouragement, and the time we spent together.

Finally, I must thank my parents and my wife Na. They have been extremely supportive and considerate for my work and in my life. I thank for their endless love and sacrifices.

vi VITA

March 15, 1980 ………………………………….……… Born – Shangqiu, Henan, China

July, 2002 ………….……………………………………………………… B.S. Pharmacy Tianjin Medical University, Tianjin, China

July, 2005 ………………………………………...... M.S. Organic Chemistry Shanghai Institute of Materia Medica Chinese Academy of Sciences, Shanghai, China

2005 – 2007 ………………………… Graduate Teaching Assistant, College of Pharmacy The Ohio State University

2007 – 2008 …………………… Graduate Teaching Assistant, Department of Chemistry The Ohio State University

2008 – 2010 …………………… Graduate Research Assistant, Department of Chemistry The Ohio State University

2010 ……………………………………………………. Gerhard L. Closs Student Award Inter-American Photochemistry Society

2010 ………………………… Henne Research Award Finalist, Department of Chemistry The Ohio State University

vii PUBLICATIONS

1. Burdzinski, G.; Zhang, Y.; Selvaraj, P.; Sliwa, M.; Platz, M. S. "Direct Observation of 1,2- Migration in the Excited States of Diazo Esters by Ultrafast Time Resolved IR Spectroscopy," J. Am. Chem. Soc , 2010 , ASAP .

2. Zhang, Y.; Wang, L.; Moss, R. A; Platz, M. S. "Ultrafast Spectroscopy of Arylchlorodiazirines: Hammett Correlations of Excited State Lifetimes," J. Am. Chem. Soc, 2009 , 131 , 16652-16653.

3. Zhang, Y.; Burdzinski, G.; Kubicki, J.; Vyas, S; Hadad, C. M; Sliva, M; Buntinx, G; Platz, M. S. "A Study of the S 1 Excited State of para -Methoxy 3-Phenyl-3- Methyl Diazirine by Ultrafast Time Resolved UV-Vis and IR Spectroscopies and Theory," J. Am. Chem. Soc. 2009 , 131 , 13784-13790.

4. Zhang, Y.; Kubicki, J.; Platz, M. S. "Ultrafast UV –vis and Infrared Spectroscopic Observation of a Singlet Vinylcarbene and the Intramolecular Cyclopropenation Reaction," J. Am. Chem. Soc. 2009 , 131 , 13602-13603 .

5. Zhang, Y.; Burdzinski, G.; Kubicki, J.; Platz, M. S. "Ultrafast Time Resolved Infrared Spectroscopy Study on the Photochemistry of N,N- Diethyldiazoacetamide. Rearrangement in the Excited State (RIES)," J. Am. Chem. Soc. 2009 , 131 , 9646-9647.

6. Kubicki, J.; Zhang, Y.; Wang, J.; Luk, H. L.; Peng, H.-L.; Vyas, S.; Platz, M. S. "Direct Observation of Acyl Excited States and Their Decay Processes by Ultrafast Time Resolved Infrared Spectroscopy," J. Am. Chem. Soc. 2009 , 131 , 4212-4213.

7. Zhang, Y.; Burdzinski, G.; Kubicki, J.; Platz, M. S. "Direct Observation of Carbene and Diazo Formation from Aryldiazirines by Ultrafast Infrared Spectroscopy," J. Am. Chem. Soc. 2008 , 130, 16134-16135.

viii 8. Zhang, Y.; Kubicki, J.; Wang, J.; Platz, M. S. "2-Naphthyl(carbomethoxy)carbene Revisited: Combination of Ultrafast UV –vis and Infrared Spectroscopic Study," J. Phys. Chem. A 2008 , 112 , 11093-11098.

9. Wang, J.; Zhang, Y.; Kubicki, J.; Platz, M. S. “Ultrafast studies of Some Diarylcarbenes” Photochem. Photobiol. Sci. 2008 , 7, 552-557.

10. Porchia, L. M.; Guerra, M.; Wang, Y.-C.; Zhang, Y.; Espinosa, A. V.; Shinohara, M.; Kulp, S. K.; Kirschner, L. S.; Saji, M.; Chen, C.-S.; Ringel, M. D. "2-Amino- N-{4-[5-(2-phenanthrenyl)-3-(trifluoromethyl)-1H-pyrazol-1-yl]-phenyl} Acetamide (OSU-03012), a Celecoxib Derivative, Directly Targets p21-Activated Kinase," Mol. Pharmacol. 2007 , 72 , 1124-1131.

11. Zhang, Y.; He, Q.; Ding, H.; Wu, X.; Xie, Y. "Improved Synthesis of Tadalafil," Org. Prep. Proc. Int. 2005 , 37 , 99-102.

FIELDS OF STUDY

Major Field: Chemistry

Physical Organic Chemistry

ix TABLE OF CONTENTS

Page

Abstract...... ii

Dedication...... iv

Acknowledgements...... v

Vita...... vii

Puplications...... viii

List of Schemes...... xvi

List of Tables ...... xviii

List of Figures...... xxii

CHAPTERS:

1. Introduction...... 1

1.1. Introduction...... 1

1.2. Traditional Product Studies in the 1950s...... 1

1.3. Matrix Isolation Spectroscopic Studies in the 1960s...... 2

1.4. Computational Chemistry in the 1970s...... 3

1.5. Nanosecond Laser Flash Photolysis in the 1980s...... 4

1.6. Ultrafast Spectroscopic Methods since the 1990s ...... 5

x 1.7. Ultrafast and Theoretical Studies of Diazirine Excited states and Carbenes...... 6

1.8. References for Chapter 1 ...... 9

2. Theory of Phenyldiazirine Excited States...... 13

2.1. Introduction...... 13

2.2. Computational Methods...... 17

2.3. Results and Discussion ...... 18

2.3.1. Phenyldiazirine ...... 18

2.3.1.1. Ground state equilibrium geometry, vertical excitations, and electronic difference density plots of phenyldiazirine...... 18

2.3.1.2. Excited State Equilibrium Geometries of Phenyldiazirine ...... 22

2.3.2. Phenyldiazomethane ...... 27

2.3.2.1. Ground state equilibrium geometry, vertical excitations, and electronic difference density plots of phenyldiazomethane...... 27

2.3.2.2. Excited State Equilibrium Geometries of Phenyldiazomethane...... 30

2.3.3. Potential Energy Surfaces Scan ...... 33

2.3.3.1 Ground State and S 1 Excited State Potential Energy Surfaces Scan of Phenyldiazirine ...... 33

2.3.3.2 Ground State and S 1 Excited State Potential Energy Surface Scans of Phenyldiazomethane ...... 38

2.4. Conclusions...... 43

2.5. References for Chapter 2 ...... 44

3. Direct Observation of Carbene and Diazo Formation from Aryldiazirines by Ultrafast Infrared Spectroscopy...... 48

3.1. Introduction...... 48

xi 3.2. Ultrafast Spectroscopic Results ...... 50

3.2.1. 3-Chloro-3-phenyldiazirine...... 50

3.2.2. Phenyldiazirine ...... 53

3.2.3. Phenylmethyldiazirine ...... 60

3.2.4. p-Biphenyldiazirine...... 63

3.2.5. p-Biphenylmethyldiazirine ...... 70

3.3. Discussion...... 72

3.3. Conclusions...... 74

3.4. Calculations...... 74

3.5. Ultrafast Spectroscopy...... 75

3.6. Synthesis ...... 75

3.6.1. Phenyldiazirine (PhCN 2H)...... 76

3.6.2. Phenylmethyldiazirine (PhCN 2CH 3)...... 77

3.6.3. p-Bihenyldiazirine ( p-BpCN 2H) ...... 78

3.6.4. p-Bihenylmethyldiazirine ( p-BpCN 2CH 3)...... 79

3.7. References for Chapter 3 ...... 79

4. A Study of the S 1 Excited State of para -Methoxy 3-Phenyl-3-Methyl Diazirine by Ultrafast Time Resolved UV–VIS and IR Spectroscopies and Theory...... 83

4.1. Introduction...... 83

4.2. Ultrafast Spectroscopic Results ...... 86

4.2.1. Ultrafast UV–vis Spectroscopy with 375 nm Excitation...... 86

4.2.2. Ultrafast IR Spectroscopy with 400 nm Excitation ...... 89

4.2.3. Ultrafast IR Spectroscopy with 270 nm Excitation ...... 97 xii

4.3. Discussions ...... 104

4.4. Conclusions...... 108

4.5. Experimental and Computational Section ...... 109

4.5.1. Materials ...... 109

4.5.2. Ultrafast Spectroscopy...... 110

4.5.2.1. Femtosecond Broadband UV–vis Transient Absorption Spectrometer. 110

4.5.2.2. Femtosecond IR Transient Absorption Spectrometer...... 110

4.5.3. Calculations...... 111

4.6. References for Chapter 4 ...... 112

5. Ultrafast Spectroscopy of Arylchlorodiazirines: Hammett Correlations of Excited State Lifetimes ...... 115

5.1. Introduction...... 115

5.2. Ultrafast Spectroscopic Results ...... 117

5.3. Computational Results...... 125

5.4. Discussion...... 129

5.4.1. Doublet Absorption Spectrum ...... 129

5.4.2. Biexponential Decay...... 130

5.4.3. Hammett Correlation ...... 131

5.4.4. Reactions of the S 1 State with ...... 133

5.5. Conclusions...... 138

5.6. Experimental Section...... 138

5.6.1. Ultrafast Spectroscopy...... 138

xiii 5.6.2. Calculations...... 139

5.6.3. Materials ...... 139

5.7. References for Chapter 5 ...... 140

6. Studies of the Substitution and Wavelength Effects on the Photochemistry of Arylalkyldiazirines by Ultrafast Time-Resolved UV–VIS Spectroscopy ...... 143

6.1. Introduction...... 143

6.2. Ultrafast UV–vis Spectroscopies Results and Discussion...... 144

6.2.1. Ultrafast LFP (350 nm) of phenyldiazirine...... 144

6.2.2. Ultrafast LFP (350 nm) of 3-methyl and other 3-alkyl-3-phenyldiazirines.. 146

6.2.3. Ultrafast LFP of p-biphenyldiazirine with 375, 350 and 275 nm Excitations ...... 150

6.2.4. Ultrafast LFP of p-biphenylmethyldiazirine with 375, 350, 310, and 270 nm Excitations...... 163

6.3. Conclusions...... 175

6.4. Ultrafast Spectroscopy...... 176

6.5. Synthesis ...... 176

6.6. References for Chapter 6 ...... 177

7. Ultrafast Time-Resolved UV–Visible and Infrared Spectroscopic Studies on Styrylcarbomethoxy Carbene...... 179

7.1. Introduction...... 179

7.2. Ultrafast Time-resolved Spectroscopic Studies...... 182

7.2.1. Ultrafast Time-resolved UV–vis Study with 1 in Acetonitrile ...... 182

7.2.2. Ultrafast Time-resolved UV–vis Study in Methanol...... 189

7.2.3. Ultrafast Time-resolved IR Studies in Chloroform ...... 193 xiv

7.2.4. Ultrafast Time-resolved IR Studies in Methanol-OD...... 205

7.3. Theoretical Studies...... 207

7.4. Conclusions...... 214

7.5. Calculations...... 214

7.6. Ultrafast Spectroscopy...... 215

7.7. Synthesis ...... 215

7.8. References for Chapter 7 ...... 217

8. Ultrafast Time-Resolved Infrared Spectroscopy Study on the Photochemistry of N,N- Dethyldiazoacetamide: Rearrangement in the Excited States (RIES) ...... 221

8.1. Introduction...... 221

8.2. Ultrafast Spectroscopic Results ...... 224

8.2.1. Ultrafast IR Studies in Chloroform...... 224

8.2.2. Ultrafast IR Studies in Methanol-OD ...... 228

8.3. Discussions and Calculations...... 236

8.4. Conclusions...... 245

8.5. Computational and Experimental Details ...... 245

8.5.1. Calculations...... 245

8.5.2. Ultrafast Spectroscopy...... 245

8.5.3. Synthesis and Materials ...... 246

8.6. References for Chapter 8 ...... 246

BIBLIOGRAPHY...... 249

APPENDIX...... 263

xv LIST OF SCHEMES

Scheme page

2.1. Two decomposition pathways of phenyldiazirine...... 15

2.2. Two mechanisms proposed in the carbene addition to ethylene...... 15

3.1. The classical chemical reactions of diazirines...... 49

3. 2. The possible reaction pathways from singlet phenylcarbene from prior studies...... 57

4.1. The proposed zwitterionic species proposed in previous studies...... 85

4.2. The proposed decay pathways from the excited states of para -methoxy-3-phenyl-3- methyl diazirine...... 108

5.1. Classical photochemical reaction pathways of arylhalodiazirines...... 116

5.2. The mechanism of rearrangement in the excited state (RIES) in the photolysis of diazirines...... 127

6.1. 1,2-Hydrogen shift in the electronically excited state of a diazirine (S 1)...... 148

7.1. The intramolecular reactions of vinylmethylenes...... 180

7.2. Proposed photochemical reactions of PhCH=CHCN2CO 2CH 3 upon photolysis.... 182

7.3. Photochemical reactions of PhCH=CHCN 2CO 2CH 3 proposed by Schmitz...... 192

8.1. Intramolecular reactions in photolytic decomposition of N,N-diethyldiazoacetamide and methyl diazoacetate in solution...... 222

8.2. The dual mechanism in β- and γ-lactam formation in the photolysis of N,N- diethyldiazoacetamide in solution...... 223

xvi 8.3. Formation of singlet carbene from the photolysis of N,N-diethyldiazoacetamide and its reaction with MeOD...... 230

8.4. Photochemical reactions for N,N-diethyldiazoacetamide (DZA) in chloroform and CH 3OD from ultrafast time-resolved studies...... 233

8.5. Formation of singlet carbene from the photolysis of N,N-diethyldiazoacetamide and its reaction with MeOD. The cation is formed by protonation of singlet carbene.. 236

xvii

LIST OF TABLES

Table Page

2.1. Vertical excitation energies, oscillator strengths, and the dominant occupied to virtual orbital configurations (>10%) contributing to the three lowest energy singlet excitations of phenyldiazirine calculated at the TD-B3LYP/6- 311+G(d,p)//B3LYP/6-31+G(d) and RI-CC2/TZVP levels of theory...... 19

2.2. The charge analysis of S 1, S 2 and S 3 states relative to S 0 state of phenyldiazirine optimized at the RI-CC2/TZVP level of theory. Charges on the ground state were subtracted to show the changes in the excited states. Positive values indicate that the atom is more positively charged than in the ground state, while negative charges indicate that atom is more negatively charged than in ground state...... 24

2.3. Dipole moments (debye), relative energies (kcal/mol), and some geometric parameters of S 0, S 1, S 2 and S 3 states of phenyldiazirine optimized at the RI- CC2/TZVP level of theory...... 26

2.4. Vibrational frequencies (cm -1) and their respective intensities (km/mol) of ground and excited states of phenyldiazirine computed at the RI-CC2/TZVP level of theory...... 26

2.5. Vertical excitation energies, oscillator strengths, and the dominant occupied to virtual orbital configurations (>10%) contributing to the three lowest energy singlet excitations of phenyldiazomethane calculated at the TD-B3LYP/6-311+G(d,p)// B3LYP/6-31+G(d) and RI-CC2/TZVP levels of theory...... 28

2.6. Calculated transition states energies and related species with various methods. Energies are in kcal/mol relative to phenyldiazomethane. ZPE is not included. Singlet point energy with the MP2 method were calculated with the geometry optimized at the B3LYP/6-311+G(d,p) level of theory. TS1 is the transition state of carbene formation from phenyldiazirine, and TS3 is the transition state of carbene formation from phenyldiazomethane...... 41

xviii

3.1. Structure of singlet phenylcarbene ( 1PhCH) optimized at the B3LYP/6-31+G(d) level of theory (frequencies scaled by 0.9614) and cartesian coordinates...... 58

3.2. Structure of triplet phenylcarbene ( 3PhCH) optimized at the B3LYP/6-31+G(d) level of theory (frequencies scaled by 0.9614) and cartesian coordinates...... 59

1 3.3. Structure of phenylmethylcarbene ( PhCCH 3) optimized at the B3LYP/6-31+G(d) level of theory (frequencies scaled by 0.9614) and cartesian coordinates...... 62

3.4. Structure of singlet p-biphenylcarbene ( 1BpCH) optimized at the B3LYP/6-31+G(d) level of theory (frequencies scaled by 0.9614) and cartesian coordinates...... 65

1 1 3.5. The lifetimes (ps) of singlet carbenes BpCH and BpCCH 3 obtained from ultrafast laser flash photolysis with both UV–vis (diazo compound) and IR detection (diazirine)...... 74

4.1. Amplitudes and lifetimes (ps) of transient species produced by ultrafast time- resolved laser flash photolysis ( λex = 375 nm) of p-methoxy-3-phenyl-3-methyl diazirine ( p-CH 3OC 6H4CN 2CH 3) in selected solvents with UV–vis detection. Lifetimes are obtained by fitting the kinetic traces to an exponential function...... 87

4.2. TD-B3LYP and RI-CC2 vertical excitation energies, oscillator strengths, and the dominant (> 10%) occupied and virtual orbitals contributing to the three lowest energy singlet excitations of p-methoxy-3-phenyl-3-methyl diazirine ( p- CH 3OC 6H4CN 2CH 3) using the optimized geometry for the S 0 state...... 92

4.3. Optimized structure of the singlet state of p-methoxy-phenylmethyl carbene ( p- CH 3OC 6H4CCH 3) at the B3LYP/6-31G(d) level of theory...... 102

4.4. The singlet-triplet energy gap (kcal/mol) of p-methoxy-phenylmethyl carbene ( p- CH 3OC 6H4CCH 3) calculated by the PCM model at the B3LYP/6-31G(d) level of theory (positive values indicate a triplet ground state and negative values indicate a singlet ground state)...... 103

5.1. Singlet phenylchlorocarbene ( 1PhCCl) vertical excitation energies computed at the B3LYP/6-311+G(d,p)//B3LYP/6-31+G(d) level of theory...... 118

5.2. Chlorophenyldiazirine (3, Y = H) vertical excitation energies computed at the B3LYP/6-311+G(d,p)//B3LYP/6-31+G(d) level of theory...... 126

xix 5.3. Lifetimes (ps) of transient absorptions of arylchlorodiazirines obtained by ultrafast laser flash photolysis with 375 nm pulses (300 fs). Only the long components of the transient decays are tabulated...... 132

6.1. Lifetimes of various transient absorptions of S 1 states, measured at maximum absorption, produced by ultrafast LFP (350 nm) of alkylphenyldiazirines...... 148

6.2. Vertical excitation energies, oscillator strengths, and the dominant occupied to virtual orbital configurations (>10%) contributing to the three lowest energy singlet excitations of biphenyldiazirine (BpCN 2H) calculated at the TD-B3LYP/6- 311+G(d,p)//B3LYP/6-31+G(d) and RI-CC2/TZVP levels of theory...... 158

6.3. Calculations of p-biphenylmethyldiazirine (BpCN 2CH 3) at the B3LYP/6- 311+G(d,p)//B3LYP/6-31+G(d) level of theory...... 168

7.1. Geometry optimization and frequency analyses for cyclopropene product (3) were obtained with the B3LYP/6-31G(d) level of theory. Frequencies were scaled by a factor of 0.9614. Vertical transition energies (nm) and oscillator strength ( f) were calculated at the TD-B3LYP/6-311+G(d,p)//B3LYP/6-31G(d) level of theory..... 198

7.2. Geometry optimization and frequency analyses for singlet carbene 1 1 PhCH=CHCCO 2CH 3 ( 2) were obtained with the B3LYP/6-31G(d) level of theory. Frequencies were scaled by a factor of 0.9614. Vertical transition energies (nm) and oscillator strength ( f) were calculated at the TD-B3LYP/6-311+G(d,p)//B3LYP/6- 31G(d) level of theory...... 202

7.3. Geometry optimization and frequency analyses for triplet carbene 3 3 PhCH=CHCCO 2CH 3 ( 2) were obtained with the B3LYP/6-31G(d) level of theory. Frequencies were scaled by a factor of 0.9614. Vertical transition energies (nm) and oscillator strength ( f) were calculated at the TD-B3LYP/6-311+G(d,p)//B3LYP/6- 31G(d) level of theory...... 203

7.4. Summary of lifetimes ( τ /ps) of transients obtained by ultrafast LFP of PhCH=CHCN 2CO 2CH 3 (1) with UV-Vis detection ( λex = 310 nm) and IR detection (λex = 270 nm) in acetonitrile (ACN), methanol, methanol-OD, chloroform, and cyclohexane (CHX)...... 207

7.5. Relative energies ∆E (0 K) calculated for singlet vinylcarbene ( 12), triplet vinylcarbene ( 32), the transition states for the formation of cyclopropene (TS1) and (TS2) from singlet vinylcarbene ( 12) and their related products. The gas phase calculations were performed at the B3LYP/6-31G(d)// B3LYP/6-31G(d) level of theory. The PCM model is used in computing the singlet point energies in

xx methanol, chloroform (CHCl 3), acetonitrile (ACN), and cyclohexane (CHX) with geometries optimized in the gas phase. Zero point energies were not included. a The relative energies (kcal/mol) are to singlet carbene ( 12)...... 211

7.6. Transition state (TS1) for the formation of cyclopropene (3) from the singlet carbene optimized at the B3LYP/6-31G(d) level of theory. Frequencies are not scaled..... 212

7.7. Transition state (TS2) for the formation of ketene from the singlet carbene optimized at the B3LYP/6-31G(d) level of theory. Frequencies are not scaled...... 213

8.1. The calculated frequencies (cm -1) for the carbonyl stretching mode in singlet carbene, β-lactam, γ-lactam, AE, cation, and DZA in the gas phase, chloroform and methanol with PCM model at the B3LYP/6-31+G(d) level of theory. Frequencies were not scaled...... 236

8.2. Optimized structure of the transition state TS1 from singlet carbene to γ-lactam at the B3LYP/6-31+G(d) level of theory...... 241

8.3. Optimized structure of the transition state TS2 from singlet carbene to β-lactam at the B3LYP/6-31+G(d) level of theory...... 242

8.4. Optimized structure of the singlet carbene at the B3LYP/6-31+G(d) level of theory...... 243

8.5. Optimized structure of the ground state N,N-diethyldiazoacetamide at the B3LYP/6- 31+G(d) level of theory...... 244

xxi

LIST OF FIGURES

Figure Page

2.1. Steady state UV–vis absorption spectrum of phenyldiazirine in pentane...... 19

2.2. Graphic representation of representative molecular orbitals of phenyldiazirine contributing to the three lowest energy singlet excitations calculated at the B3LYP/6-311+G(d,p) level of theory...... 20

2.3. Excited state difference density plots for the S 1, S 2 and S 3 states relative to the ground state electron density for phenyldiazirine, as calculated at the RI-CC2/TZVP level of theory. A red surface surrounds areas where electron density is depleted after vertical excitation from S0; a green surface surrounds areas where electron density is accumulated in the excited state...... 21

2.4. Optimized geometries for the ground and electronic excited states of phenyldiazirine at the RI-CC2/TZVP level of theory. Distances are shown in angstroms...... 23

2.5. Graphic representation of representative molecular orbitals of phenyldiazomethane contributing to the three lowest energy singlet excitations calculated at the B3LYP/6-311+G(d,p) level of theory...... 28

2.6. Excited state difference density plots for the S 1, S 2 and S 3 states relative to the ground state electron density for phenyldiazomethane, as calculated at the RI- CC2/TZVP level of theory. A red surface surrounds areas where electron density is depleted after vertical excitation from S 0; a green surface surrounds areas where electron density is accumulated in the excited state...... 29

2.7. Optimized geometries of the ground and excited states of phenyldiazomethane at the RI-CC2/TZVP level of theory. Bond lengths are in Angstroms, and bond angles are in degrees...... 31

xxii 2.8. The composite energy diagram of the vertical and adiabatic excited states of phenyldiazirine (left) and phenyldiazomethane (right), as calculated at the RI- CC2/TZVP level of theory. The energies are in kcal/mol and are relative to the S 0 ground state of phenyldiazomethane...... 32

2.9. The relaxed optimized geometries of phenyldiazirine after extension of the C 12 –N15 bond at the RI-CC2/TZVP level of theory, starting from the ground (S 0, left panel) and S 1 excited (right panel) state optimized geometries. The transition structures of TS1 (ground state) and TS2 (S 1 excited state) obtained are shown on top. Energies are relative to the ground state equilibrium geometry of phenyldiazirine in kcal/mol. Bond distances are in Angstroms. Bond angles are in degrees...... 35

2.10. The relaxed optimization of phenyldiazirine by fixing one C–N bond (black curve) and two C–N bonds (red curve) on the ground (S 0), S 1, and S 2 excited states potential energy surfaces at the RI-CC2/TZVP level of theory. Energies are relative to the ground state equilibrium geometry of phenyldiazirine in kcal/mol. Bond distances are in Angstroms...... 36

2.11. The relaxed optimized geometries of phenyldiazomethane after extension of the C12 –N15 bond at the RI-CC2/TZVP level of theory, starting from the S 0 ground state and leading to the TS3 transition state for extrusion of molecular . Energies are relative to the ground state equilibrium geometry in kcal/mol. Bond distances are in Angstroms. Bond angles are in degrees. Atom labeling is provided in TS3 structure to the right...... 39

2.12. The contour plot of the potential energy surface of phenyldiazirine along the two C– N bond coordinates, as computed at the B3LYP/6-31+G(d) level of theory...... 40

2.13. The thermal transition states for the formation of phenylcarbene from phenyldiazirine (TS1) and phenyldiazomethane (TS3) calculated at the RI- CC2/TZVP level of theory on the ground state surfaces. Bond lengths are in Angstroms...... 42

3.1. Transient IR spectra produced by photolysis of chlorophenyldiazirine ( λex = 270 nm) in chloroform. (a) The diazo band in the region 2060 – 1940 cm -1 with time delays of 3 – 236 ps. (b) The carbene band in the region 1630 – 1550 cm -1 with time delays of 4 – 164 ps...... 51

3.2. Transient IR spectra produced by photolysis of chlorophenyldiazirine ( λex = 270 nm) in chloroform. The integration of the diazo band (2060 – 1940 cm -1) versus time delay (note the time delays in logarithmic scale)...... 53

xxiii 3.3. Transient IR spectra produced by photolysis of phenyldiazirine ( λex = 270 nm) in chloroform. The diazo band in the spectral region of 2100 – 2035 cm -1 in a time window of 2 – 2200 ps...... 54

3.4. Transient IR spectra produced by ultrafast LFP of phenyldiazirine ( λex = 270 nm) in chloroform. (a) The formation of singlet carbene 1PhCH in a time window of 1 – 50 ps. (b) The decay of singlet carbene 1PhCH band at 1582 cm -1 in a time window of 50 – 1200 ps...... 55

3.5. The decay trace of phenylcarbene 1PhCH band monitored at 1582 cm -1 produced by ultrafast LFP of phenyldiazirine ( λex = 270 nm) in chloroform and was fit to an exponential function...... 55

3.6. The transient spectra were generated by ultrafast LFP (270 nm) of phenylmethyldiazirine in chloroform. (a) The diazo band detected in the 2060 – 1930 cm -1 spectral widow with time delays 1 – 100 ps. (b) The carbene band detected in the 1640 – 1550 cm -1 spectral widow with time delays of 2 – 30 ps ..... 61

3.7. Ultrafast LFP (λex = 270 nm) of p-biphenyldiazirine in acetonitrile. (a) Transient spectra of diazo compound in a time window of 2 – 180 ps. (b) The integration of the diazo band intensities versus time delays...... 63

3.8. Transient spectra were generated by ultrafast LFP (270 nm) of p-biphenyldiazirine in chloroform. (a) Transient IR spectra showing the growth of 1BpCH carbene in a time window of 4 – 65 ps. (b) Transient IR spectra showing the decay of 1BpCCH carbene in a time window of 65 – 1500 ps...... 64

3.9. Ultrafast LFP ( λex = 270 nm) of p-biphenyldiazirine in chloroform. (a) The kinetic trace of 1BpCH carbene band monitored at 1570 cm -1 by fitting to an exponential function. (b) The kinetic trace of 1BpCH carbene band obtained by intensity integration...... 66

1 3.10. The transient IR spectra of p-biphenylcarbene BpCH obtained by ultrafast LFP ( λex = 270 nm) of p-biphenyldiazirine ( p-BpCN 2H) in CH 2Cl 2...... 68

1 3.11. The transient IR spectra of p-biphenylcarbene BpCH obtained by ultrafast LFP ( λex = 270 nm) of p-biphenyldiazirine ( p-BpCN 2H) in cyclohexene (CHE)...... 68

1 3.12. The transient IR spectra of p-biphenylcarbene BpCH obtained by ultrafast LFP ( λex = 270 nm) of p-biphenyldiazirine ( p-BpCN 2H) in methanol-O-d (MeOD)...... 69

1 3.13. The lifetimes of p-biphenylcarbene BpCH obtained by ultrafast LFP ( λex = 270 nm) of p-biphenyldiazirine ( p-BpCN 2H) in (a) CH 2Cl 2, (b) cyclohexene (CHE), and (c)

xxiv methanol-O-d (MeOD). The decay traces were probed at wavenumbers of maximum absorption and fitted to the exponential function A = A exp(-t/τ) + y0...... 69

3.14. Transient spectra generated by ultrafast LFP ( λex = 270 nm) of p- biphenylmethyldiazirine in chloroform. (a) Transient IR spectra showing the growth 1 of BpCCH 3 carbene within time windows 2 – 65 ps. (b) Transient IR spectra 1 showing the decay of BpCCH 3 carbene within time windows 65 – 1763 ps...... 70

3.15. Ultrafast LFP ( λex = 270 nm) of p-biphenylmethyldiazirine (BpCN2CH 3) in 1 chloroform. The decay trace of the p-biphenylmethylcarbene BpCCH 3 band monitored at 1585 cm -1 was fitted to an exponential function...... 71

1 1 1 3.16. Predicted vibrational bands of singlet carbenes PhCH, PhCCH 3, BpCH and 1 BpCCH 3 at the B3LYP/6-31+G(d) level of theory...... 72

4.1. Transient spectra produced by ultrafast laser flash photolysis ( λex = 375 nm) of p- methoxy-3-phenyl-3-methyl diazirine ( p-CH 3OC 6H4CN 2CH 3) in acetonitrile with a time window of 5–800 ps...... 86

4.2. Ultrafast time-resolved decay curves with UV–vis detection obtained from p- methoxy-3-phenyl-3-methyl diazirine ( p-CH 3OC 6H4CN 2CH 3) with 375 nm excitation. Kinetic traces decay probed at maximum absorbance in (a) cyclohexane, (b) acetonitrile, (c) methanol, (d) 2,2,2-trifluoroethanol, (e) MeOD, and (f) CHCl 3...... 88

4.3. Transient UV–vis spectra produced by ultrafast photolysis of p-methoxy-3-phenyl- 3-methyl diazirine ( p-CH 3OC 6H4CN 2CH 3). The spectra were generated by ultrafast LFP ( λex = 375 nm) in (a) cyclohexane, (b) methanol, and (c) 2,2,2-trifluoroethanol...... 89 4.4. Transient IR spectra produced by ultrafast photolysis of p-methoxy-3-phenyl-3- methyl diazirine ( p-CH 3OC 6H4CN 2CH 3) in chloroform. The spectra were generated by ultrafast LFP ( λex = 400 nm) with a time window of 30–1600 ps. The inset shows the kinetics of decay at 1580 cm -1 by fitting to an exponential function...... 90

4.5. Ultrafast LFP ( λex = 400 nm) of p-methoxy-3-phenyl-3-methyl diazirine ( p- CH 3OC 6H4CN 2CH 3). Kinetic traces probed at maximum absorbance in (a) methanol-d4 (CD 3OD) and (b) cyclohexane fitted to an exponential function...... 91

4.6. Steady state UV–vis absorption spectra of p-methoxy-3-phenyl-3-methyl diazirine (p-CH 3OC 6H4CN 2CH 3) in pentane...... 92

4.7. The optimized geometries of S 0 (left) and S 1 (right) states of p-CH 3OC 6H4CN 2CH 3 at RI-CC2/TZVP level of theory (the bond lengths are in Angstroms)...... 93

xxv

4.8. Frequencies of the phenyldiazirine electronic excited states (S 1, S 2 and S 3) and ground state S 0 optimized at the RI-CC2/TZVP level of theory...... 94

4.9. Dynamics of solvation of the S 1 state of p-methoxy-3-phenyl-3-methyl diazirine ( p- CH3OC 6H4CN 2CH 3) by ultrafast IR ( λex = 400 nm). (a) cyclohexane; (b) chloroform; (c) methanol-d4...... 95

4.10. Transient IR spectra produced by ultrafast photolysis of p-methoxy-3-phenyl-3- methyl diazirine in chloroform. The spectra were generated by ultrafast LFP ( λex = 400 nm) with a time window of 30–1600 ps. The inset shows the signal growth at 2030 cm -1 by fitted to an exponential function...... 97

4.11. Transient IR spectra produced by ultrafast photolysis of p-methoxy-3-phenyl-3- methyl diazirine in chloroform. (a) The spectra were generated by ultrafast LFP ( λex = 270 nm) with a time delays of 1 – 200 ps. (b) The growth of the diazo band obtained by fitting the band integration to a biexponential function with the slow component fixed at 240 ps...... 98

4.12. Transient IR spectra produced by ultrafast photolysis of p-methoxy-3-phenyl-3- methyl diazirine in chloroform. The spectra were generated by ultrafast LFP ( λex = 270 nm) with a time window of (a) 1.5 – 54 ps, and (b) 54 – 3047 ps...... 99

4.13. Ultrafast IR spectroscopy performed on p-methoxy-3-phenyl-3-methyl diazirine ( p- CH 3OC 6H4CN 2CH 3) in chloroform with 270 nm excitation. (a) Kinetics of carbene decay probed at 1584 cm -1. (b) Kinetics of formation of carbene by fitting the band integrals vs time delay into an exponential function...... 100

4.14. Ultrafast IR spectroscopy of p-methoxy-3-phenyl-3-methyl diazirine ( p- CH 3OC 6H4CN 2CH 3) in methanol-O-d with 270 nm excitation. A kinetic trace of carbene decay was probed at 1584 cm -1 by fitting into a biexponential function with the slow component fixed at 390 ps...... 103

4.15. Laser flash photolysis ( λex = 355 nm and 308 nm) study of p-methoxy-3-phenyl-3- methyl diazirine ( p-CH 3OC 6H4CN 2CH 3) in chloroform. Peak absorbance of carbene- pyridine ylide at different pyridine concentration...... 106

5.1. Ultrafast laser flash photolysis ( λex = 375 nm) of phenylchlorodiazirine (3a, Y = H) in ACN. (a) Transient spectra of 3a with a time window of 0.3 – 300 ps. (b) Kinetic traces of transient absorptions probed at 640 nm and fitted to exponential functions...... 117

xxvi 5.2. Ultrafast laser flash photolysis ( λex = 375 nm) of arylchlorodiazirine (3b, Y = p-CH 3) in ACN. (a) Transient spectra of 3b with a time window of 3 – 425 ps. (b) Kinetic traces of transient absorptions were probed at maximum absorptions and fitted to an exponential function...... 120

5.3. Ultrafast laser flash photolysis ( λex = 375 nm) of arylchlorodiazirine (3c, Y = p- OCH 3) in ACN. (a) Transient spectra of 3c with a time window of 27 – 2194 ps. (b) Kinetic traces of transient absorptions were probed at maximum absorptions and fitted to an exponential function...... 120

5.4. Ultrafast laser flash photolysis ( λex = 375 nm) of arylchlorodiazirine (3d, Y = p-Cl) in ACN. (a) Transient spectra of 3d with a time window of 1 – 425 ps. (b) Kinetic traces of transient absorptions probed at maximum absorptions and fitted to exponential functions...... 121

5.5. Ultrafast laser flash photolysis ( λex = 375 nm) of arylchlorodiazirine (3e, Y = m-Cl) in ACN. (a) Transient spectra of 3e with a time window of 1 – 98 ps. (b) Kinetic traces of transient absorptions probed at maximum absorptions and fitted to an exponential function...... 121

5.6. Ultrafast laser flash photolysis ( λex = 375 nm) of arylchlorodiazirine (3f, Y = p-CF 3) in ACN. (a) Transient spectra of 3f produced with a time window of 1 – 33 ps. (b) Kinetic traces of transient absorptions probed at maximum absorptions and fitted to an exponential function...... 122

5.7. Ultrafast laser flash photolysis ( λex = 375 nm) of arylchlorodiazirine in chloroform . Kinetic traces of transient absorptions probed at maximum absorption and fitted to exponential functions. (a) 3a, Y = H. (b) 3b, Y = p-CH 3. (c) 3c, Y = p-CH 3O. (d) 3d, Y = p-Cl. (e) 3e, Y = m-Cl. (f) 3f, Y = p-CF 3...... 123

5.8. Ultrafast kinetic traces ( λex = 375 nm) of arylchlorodiazirine in cyclohexane . Kinetic traces of transient absorption probed at maximum absorptions and fitted to exponential functions. (a) 3a, Y = H. (b) 3b, Y = p-CH 3. (c) 3c, Y = p-CH 3O. (d) 3d, Y = p-Cl. (e) 3e, Y = m-Cl. (f) 3f, Y = p-CF 3...... 124

5.9. S0 (left), S 1 (middle), and S 2 (right) of 3* (Y=H) optimized at the RI-CC2/TZVP level of theory; bond lengths are in angstroms...... 128

+ + 5.10. Hammett correlations for S 1 of 3*: log τ vs σp ; see Table 5.3 for τ and σp . ρ = - 1.37 ( r = -0.998) for ACN (black), ρ = -1.43 ( r = -0.995) for CHCl 3 (blue), and ρ = - 1.27 ( r = -0.996) for CHX (red)...... 133

5.11. Ultrafast time resolved kinetic traces ( λex = 375 nm) of arylchlorodiazirines in methanol . Kinetic traces of transient absorptions probed at maximum absorptions xxvii and fitted to exponential functions. (a) 3a, Y = H. (b) 3b, Y = p-CH 3. (c) 3c, Y = p- CH 3O. (d) 3d, Y = p-Cl. (e) 3e, Y = m-Cl. (f) 3f, Y = p-CF 3...... 135

5.12. Ultrafast time resolved kinetic traces ( λex = 375 nm) of arylchlorodiazirine in trifluoroethanol . Kinetic traces of transient absorptions probed at maximum absorptions and fitted to exponential functions. (a) 3a, Y = H. (b) 3b, Y = p-CH 3. (c) 3c, Y = p-CH 3O. (d) 3d, Y = p-Cl. (e) 3e, Y = m-Cl. (f) 3f, Y = p-CF 3...... 136

5.13. Ultrafast laser flash photolysis ( λex = 375 nm) of arylchlorodiazirine (3c, Y = p- CH 3O) in methanol -OD. Kinetic traces probed at maximum absorption wavelengths fitted to exponential functions...... 137

+ + 5.14. Hammett correlations for S 1 of 3*: log τ vs σp ; see Table 5.3 for τ and σp . ρ = - 1.17 ( r = -0.933) for MeOH (green), and ρ = -1.76 ( r = -0.936) for TFE (red). .... 137

6.1. Transient spectra generated by ultrafast LFP (350 nm) of phenyldiazirine in (a) acetonitrile and (b) cyclohexane, at selected time delays...... 145

6.2. Kinetic traces obtained upon monitoring at 550 nm of the transient spectra generated by ultrafast LFP (350 nm) of phenyldiazirine in (a) acetonitrile and (b) cyclohexane. The kinetic traces were fitted to a bi-exponential function with an instrument response function (IRF) of 300 fs...... 145

6.3. Transient spectra generated by ultrafast LFP (350 nm) of 3-phenyl-3- methyldiazirine in (a) acetonitrile and (b) cyclohexane, at selected time delays. .. 146

6.4. Kinetic traces obtained at 550 nm of transient spectra generated by ultrafast LFP (350 nm) of 3-methyl-3-phenyl diazirine in (a) acetonitrile and (b) cyclohexane. The kinetic traces were fitted to a bi-exponential function with an instrument response function (IRF) of 300 fs...... 147

6.5. Kinetic traces obtained at 550 nm of transient spectra generated by ultrafast LFP (350 nm) of 3-ethyl-3-phenyl diazirine in (a) acetonitrile and (b) cyclohexane. The kinetic traces were fitted to a bi-exponential function with an instrument response function (IRF) of 300 fs...... 149

6.6. Kinetic traces obtained at 550 nm of transient spectra generated by ultrafast LFP (350 nm) of 3-iso -propyl-3-phenyl diazirine (PhCN 2-i-Pr) in (a) acetonitrile and (b) cyclohexane. The kinetic traces were fitted to a bi-exponential function with an instrument response function (IRF) of 300 fs...... 149

6.7. Kinetic traces obtained at 550 nm of transient spectra generated by ultrafast LFP (350 nm) of 3-tert -butyl-3-phenyl diazirine (PhCN 2-t-Bu) in (a) acetonitrile and (b)

xxviii cyclohexane. The kinetic traces were fitted to a bi-exponential function with an instrument response function (IRF) of 300 fs...... 150

6.8. Kinetic traces obtained at 550 nm of transient spectra generated by ultrafast LFP (350 nm) of 3-trideuteromethyl-3-phenyl diazirine (PhCN 2CD 3) in (a) acetonitrile and (b) cyclohexane. The kinetic traces were fitted to a bi-exponential function with an instrument response function (IRF) of 300 fs...... 150

6.9. Transient spectra generated by ultrafast LFP (375 nm) of p-biphenyldiazirine in acetonitrile at selected time delays...... 151

6.10. Kinetic traces monitored at 425 nm of transient spectra generated by ultrafast LFP (375 nm) of p-biphenyldiazirine in (a) acetonitrile and (b) 2,2,2-trifluoroethanol and fitted to an exponential function...... 151

6.11. Transient spectra generated by ultrafast LFP (350 nm) of p-biphenyldiazirine in (a) acetonitrile (b) cyclohexane and (c) methanol at selected time delays...... 153

6.12. Kinetic traces monitored at 400 nm of transient spectra generated by ultrafast LFP (350 nm) of p-biphenyldiazirine in (a) acetonitrile (b) cyclohexane and (c) methanol at selected time delays and fitted to an exponential function...... 154

6.13. Transient spectra generated by ultrafast LFP (275 nm) of p-biphenyldiazirine in acetonitrile at selected time delays...... 155

6.14. Kinetic traces monitored at (a) 490 nm, (b) 430 nm and (c) 360 nm of transient spectra generated by ultrafast LFP (275 nm) of p-biphenyldiazirine in acetonitrile...... 156

6.15. Transient spectra generated by ultrafast LFP (275 nm) of p-biphenyldiazirine in acetonitrile at selected time delays in the spectral window of 400 – 680 nm...... 159

6.16. Transient spectra generated by ultrafast LFP (275 nm) of (a) bleached solution of p- biphenyldiazirine and (b) p-biphenylcarboxaldehyde ( p-BpCHO) in acetonitrile at selected time delays...... 161

6.17. Kinetic traces monitored at (a) 490 nm, (b) 420 nm and (c) 360 nm of the transient spectra generated by ultrafast LFP (270 nm) of p-biphenyldiazirine in methanol. 162

6.18. Kinetic traces monitored at (a) 490 nm, (b) 420 nm and (c) 360 nm of transient spectra generated by ultrafast LFP (270 nm) of p-biphenyldiazirine in cyclohexane...... 163

6.19. Transient spectra generated by ultrafast LFP (375 nm) of p-biphenylmethyldiazirine in acetonitrile at selected time delays...... 164 xxix

6.20. Kinetic traces monitored at 410 nm of transient spectra generated by ultrafast LFP (375 nm) of p-biphenylmethyldiazirine in (a) acetonitrile and (b) methanol...... 165

6.21. Transient spectra generated by ultrafast LFP (350 nm) of p-biphenylmethyldiazirine in acetonitrile at selected time delays...... 166

6.22. Kinetic traces monitored at (a) 490 nm and (b) 400 nm of transient spectra generated by ultrafast LFP (350 nm) of p-biphenylmethyldiazirine in acetonitrile...... 167

6.23. Transient spectra generated by ultrafast LFP (275 nm) of p-biphenylmethyldiazirine in acetonitrile at selected time delays...... 170

6.24. Kinetic traces monitored at (a) 490 nm, (b) 415 nm and (c) 360 nm of transient spectra generated by ultrafast LFP (275 nm) of p-biphenylmethyldiazirine in acetonitrile...... 171

6.25. Transient spectra generated by ultrafast LFP (275 nm) of (a) a bleached solution of p-biphenylmethyldiazirine and (b) BpCOCH 3 in acetonitrile at selected time delays...... 172

6.26. Kinetic traces monitored at (a) 495 nm, (b) 413 nm and (c) 360 nm of transient spectra generated by ultrafast LFP (275 nm) of p-biphenylmethyldiazirine in cyclohexane...... 173

6.27. Excited state difference density plots for the S 1, S 2 and S 3 states relative to the ground state electron density for p-biphenyldiazirine, as calculated at the RI- CC2/TZVP level of theory. A red surface surrounds areas where electron density is depleted after vertical excitation from S 0; a green surface surrounds areas where electron density is accumulated in the excited state...... 175

7.1. Transient UV–vis spectra generated by ultrafast time-resolved LFP ( λex = 310 nm) of PhCH=CHCN 2CO 2CH 3 in acetonitrile within time windows (a) 0.5 – 1.0 ps and (b) 2 – 1300 ps...... 183

1 1 7.2. Simulated absorption spectrum of singlet carbene PhCH=CHCCO 2CH 3 ( 2) in the gas phase. The vertical transition energies (nm) and oscillator strength ( f) are shown as lines from the TD-B3LYP/6-311+G(d,p)//B3LYP/6-31G(d) calculatins. The overall absorption bands are obtained by Lorentzian broadening...... 184

3 3 7.3. Simulated absorption spectrum of triplet carbene PhCH=CHCCO 2CH 3 ( 2) in the gas phase. The vertical transition energies (nm) and oscillator strength ( f) are shown as lines from the TD-B3LYP/6-311+G(d,p)//B3LYP/6-31G(d) calculations. The overall absorption bands are obtained by Lorentzian broadening...... 184 xxx

7.4. The kinetic traces monitored at 390 nm obtained by ultrafast time-resolved LFP ( λex = 310 nm) of PhCH=CHCN 2CO 2CH 3 (1) in (a) acetonitrile, (b) cyclohexane, and (c) chloroform. The kinetic traces were fitted to exponential functions...... 186

7.5. Simulated absorption spectrum of cyclopropene (3) in the gas phase. The vertical transition energies (nm) and oscillator strength (f) are shown as lines from the TD- B3LYP/6-311+G(d,p)//B3LYP/6-31G(d) calculations. The overall absorption bands are obtained by Lorentzian broadening...... 187

7.6. Transient UV–vis spectra generated by ultrafast time-resolved LFP ( λex = 310 nm) of PhCH=CHCN 2CO 2CH 3 in chloroform within time windows (a) 0.3 – 4.8 ps and (b) 5 – 600 ps...... 187

7.7. Transient spectra produced by LFP ( λex = 310 nm) of bleached solution of PhCHCHCN 2CO 2CH 3 (1) in acetonitrile...... 189

7.8. Transient UV–vis spectra generated by ultrafast time-resolved LFP ( λex = 310 nm) of PhCHCHCN 2CO 2CH 3 in methanol in time windows (a) 1.3 – 3.2 ps and (b) 9 – 118 ps...... 190

7.9. Transient UV–vis spectra generated by ultrafast time-resolved LFP ( λex = 310 nm) of PhCHCHCN 2CO 2CH 3 in methanol-OD in time windows (a) 2 – 13 ps and (b) 22 – 353 ps...... 191

7.10. The kinetic traces monitored at 390 nm obtained by ultrafast time-resolved LFP of PhCHCHCN 2CO 2CH 3 (1) with 310 nm light in (a) methanol and (b) methanol-OD. The kinetic traces were fitted to a single exponential function...... 191

7.11. Simulated absorption spectrum of cation in the gas phase formed by protonation of the singlet carbene. The vertical transition energies (nm) and oscillator strength ( f) are shown as lines from the TD-B3LYP/6-311+G(d,p)//B3LYP/6-31G(d) calculations. The overall absorption bands are obtained by Lorentzian broadening...... 193

7.12. Transient IR spectra produced by ultrafast time-resolved LFP ( λex = 270 nm) of -1 PhCHCHCN 2CO 2CH 3 in chloroform with a spectral window of 2160 – 2000 cm . The dotted curve is scaled conventional FTIR spectra in chloroform...... 194

7.13. The kinetic trace of the diazo band obtained by ultrafast time-resolved LFP ( λex = 270 nm) of PhCHCHCN 2CO 2CH 3 in chloroform. The kinetic trace is monitored at 2095 cm -1 and is fitted to an exponential function...... 195

xxxi 7.14. Transient IR spectra produced by ultrafast time-resolved LFP ( λex = 270 nm) of - PhCHCHCN 2CO 2CH 3 (1) in chloroform with a spectral window of 1750 – 1640 cm 1. The dotted curve is the stationary FTIR spectrum in chloroform...... 195

7.15. Transient IR spectra produced by ultrafast time-resolved LFP ( λex = 270 nm) of - PhCHCHCN 2CO 2CH 3 (1) in chloroform with a spectral window of 1790 – 1680 cm 1. The dotted curve is the stationary FTIR spectra in chloroform...... 199

7.16. Transient spectra at selected time delays obtained by subtracting the bleaching band of precursor from the transient spectra produced by ultrafast photolysis (270 nm) of -1 PhCHCHCN 2CO 2CH 3 in chloroform with a spectral window of 1790 – 1680 cm ...... 200

7.17. Stationary FTIR spectra of PhCHCHCN 2CO 2CH 3 (1) in chloroform taken before and after laser flash photolysis (270 nm)...... 200

7.18. Band intensity integration of the transient spectra of cyclopropene (1) obtained from ultrafast LFP ( λex = 270 nm) of PhCHCHCN 2CO 2CH 3 in chloroform with a spectral window of 1790 – 1680 cm -1...... 201

7.19. Transient IR spectra produced by ultrafast time-resolved LFP ( λex = 270 nm) of - PhCHCHCN 2CO 2CH 3 (1) in chloroform with a spectral window of 1800 – 1560 cm 1. The transient spectra were obtained by combining the data of multiple experiments with different spectral windows. The dotted curve is the stationary FTIR spectra in chloroform. The green bars are vibrational frequencies of cyclopropene (3) and singlet ( 12) and triplet ( 32) carbenes calculated at the B3LYP/6-31G(d) level of theory and scaled using the experimental (FTIR) frequencies of diazo precursor (1). The calculated intensities were scaled...... 204

7.20. Transient IR spectra produced by ultrafast time-resolved LFP ( λex = 270 nm) of PhCHCHCN 2CO 2CH 3 (1) in methanol-OD with a spectral window of 1750 – 1630 cm -1...... 206

7.21. Kinetic traces monitored at 1684, 1648, and 1710 cm -1 of transient IR spectra obtained by ultrafast LFP ( λex = 270 nm) of PhCHCHCN 2CO 2CH 3 (1) in methanol- OD. Fitting of the kinetic traces at 1648 and 1710 cm -1 produces the same time -1 constants of τ1 = 7.0 ± 0.1 ps and τ2 = 95 ± 12 ps. The flat trace at 1648 cm indicates the isosbestic point...... 206

7.22. Selected structural parameters calculated for singlet and triplet states of vinylmethylene, vinylchlorocarbene, and styrylcarbomethoxycarbene (2) at the B3LYP/6-31G(d) level of theory in gas phase. Bond lengths in angstroms. Bond angles in degrees...... 208

xxxii 7.23. The reaction paths predicted for the formation of cyclopropene (TS1) and ketene 1 (TS2) from singlet vinylcarbene PhCHCHCCO 2CH 3. Relative free energies G (298 K) were calculated at the B3LYP/6-31G(d) level of theory in the gas phase. 210

8.1. Conventional photolysis of N,N-diethyldiazoacetamide (DZA) with 270 nm laser pulses in chloroform. Stationary FTIR spectra were taken before and after photolysis of a fresh solution in chloroform. The FT-IR spectrum of an authentic sample of γ- lactam is recorded in chloroform. Vibrational frequencies for carbonyl stretches of DZA, the β- and γ-lactams, and singlet carbene in chloroform were obtained from calculations with B3LYP/6-31+G(d) using the PCM model (cf. Table 8. 1). The calculated frequencies were scaled by 0.978. The scaling factor was obtained by calibrating the predicted carbonyl frequency of DZA with its experimental value.225

8.2. Transient IR spectra were generated by ultrafast LFP (270 nm) of N,N- diethyldiazoacetamide in chloroform over time windows of (a) 2 – 26 ps and (b) 26 – 1020 ps. The dashed curves are scaled FTIR spectra recorded for an authentic sample of γ-lactam in a chloroform solution...... 226

8.3. Kinetic traces obtained by ultrafast IR spectroscopy (270 nm) of N,N- diethyldiazoacetamide in chloroform. (a) The kinetic trace monitored at 1745 cm -1 for β-lactam was fitted to an exponential function. (b) The kinetic trace monitored at 1669 cm -1 for γ-lactam was fitted to an exponential function...... 227

8.4. Transient IR spectra were generated by ultrafast LFP (270 nm) of N,N- diethyldiazoacetamide (DZA) in chloroform in time windows of 0.7 – 94 ps. The dashed curves are scaled FTIR spectra of γ-lactam and DZA in chloroform...... 228

8.5. Conventional photolysis (270 nm) of N,N-diethyldiazoacetamide (DZA) in CH 3OD. Stationary FT-IR spectra were taken before and after photolysis of a fresh solution in CH 3OD. The FTIR spectra of AE and authentic γ-lactam were recorded with CH 3OD as the solvent. Vibrational frequencies for carbonyl stretches of DZA, the β- and γ-lactams, amide ether (AE), and singlet carbene in methanol were obtained from calculations with B3LYP/6-31+G(d) using the PCM model (cf. Table 8. 1). The calculated frequencies were scaled by 0.978. The scaling factor was obtained by calibrating the predicted carbonyl frequency of DZA with its experimental value.229

8.6. Transient IR spectra were generated by ultrafast LFP (270 nm) of N,N- diethyldiazoacetamide (DZA) in CH 3OD over time windows of (a) 0.7 – 31 ps and (b) 31 – 3050 ps. The dashed curves are FTIR spectra of DZA, AE, and authentic γ- lactam in MeOD...... 231

xxxiii 8.7. A contour plot (2D) of transient IR spectra generated by ultrafast LFP (270 nm) of N,N-diethyldiazoacetamide in CH 3OD in a time window of 0.7 – 3050 ps and spectral window of 1760 – 1580 cm -1...... 232

8.8. Kinetic traces produced upon ultrafast LFP (270 nm) of N,N-diethyl diazoacetamide -1 (DZA) in CH 3OD. (a) The kinetic trace monitored at 1733 cm was fitted to an exponential function. (b) The kinetic trace monitored at 1665 cm -1...... 233

8.9. Kinetic traces monitored at 1706 cm -1 produced from ultrafast LFP (270 nm) of N,N- diethyl diazoacetamide in methanol-OD and was fitted to an exponential function in the time windows of (a) 0.6 – 20 ps and (b) 20 – 3000 ps...... 234

8.10. Kinetic traces monitored at 1643 cm -1 produced from ultrafast IR spectroscopy (270 nm) of N,N-diethyl diazoacetamide in methanol-OD. The kinetic trace at 1643 cm -1 fitted to an exponential function...... 235

8.11. The transition states for the formation of γ-lactam (TS1) and β-lactam (TS2) from singlet carbene calculated at the B3LYP/6-31+G(d) level of theory. Free energies (∆G) are in kcal/mol at 298 K in the gas phase. ZPE was corrected...... 239

xxxiv CHAPTER 1

INTRODUCTION

1.1. Introduction

According to the IUPAC Gold Book , carbenes are defined as a class of organic

molecules containing a divalent carbon atom which has six electrons in its valence

1,2 shell. In the nineteenth century there were a few isolated attempts to generate the CH 2 molecule using the conventional methods of the time, 3-6 during an era when the tetravalency of carbon was not fully understood. Most chemists 7 credit the origins of

carbene compounds in their modern meaning to Buckner (in 1903) 8 and Staudinger (in

1911) 9 in regard to their studies of diazo compounds. However, information concerning carbenes in the first half of the 20 th century was mostly speculative. Systematic studies of carbenes began in the 1950s. In fact the term carbene was conceived by Doering,

Winstein, and Woodward in a famous nocturnal taxi ride in Chicago in 1950. 10 Since then, the term carbene has come into the lexicon of organic chemists and has been subject to vigorous studies with the latest techniques of each generation of chemists.

1.2. Traditional Product Studies in the 1950s

Chemical analyses of the products formed in carbene processes are painstaking

but nonetheless are still very important in studies of reactive intermediates. Chemists

1 initially understood the reactivities of carbenes through piecing together the information deduced from stable products. Hine proved the immediacy of (CCl 2) in the basic hydrolysis of chloroform using kinetics. 11 Doering and Hoffmann discovered the characteristic carbene addition reaction to alkenes to form cyclopropanes. 12 Doering et al.10,13 also discovered the indiscriminate C–H insertion reactions of with

saturated hydrocarbons.

Herzberg recorded the first electronic spectra of triplet methylene in the gas

phase 14 and first recognized the existence of singlet and triplet carbenes. Herzberg was awarded the Nobel Prize in chemistry (1971) for his contributions to the knowledge of free radicals (note that triplet methylene can be considered a radical species). 15

Upon correlation of carbene reactivities with spin multiplicities, and chemical

analyses of the stereochemistry of their reactions, chemists learned that singlet and triplet

carbenes have very different reactivities. The singlet carbene reacts with olefins in a

concerted manner to produce cyclopropanes with the stereochemistry of the alkene

retained. In contrast, the triplet reaction is stepwise and the cyclopropane products are

stereo random. These reactivity patterns are widely used as the criteria to distinguish the

spin states of carbenes and have become commonly known as the Skell-Woodworth

rule .16

1.3. Matrix Isolation Spectroscopic Studies in the 1960s

In the 1960s chemists searched for ways to directly detect carbenes, and turned to the matrix isolation technique soon after its invention. 17-19 In this approach, carbenes were isolated in cryogenic matrixes (4 – 77 K) and subject to conventional spectroscopic characterization. Bimolecular and many unimolecular reactions of carbenes are impeded

2 and carbenes are often persistent species under these conditions. Matrix isolated

20,21 22 22 diphenylcarbene (Ph 2C:), biphenylcarbene, and phenylcarbene were studied by

Wasserman, 20,22 Trozzolo, 20-24 and Closs 25,26 in their ground triplet states. Electron spin resonance (ESR) spectra of triplet species have unique features and the matrix-ESR method was enormously helpful in carbene studies.

Diazirines, the cyclic isomers of diazo compounds, also were discovered in the early 1960s. 27-29 Because of spin conservation, singlet carbenes are initially formed from

an excited singlet state of a diazo or diazirine precursor upon photolysis, and rapidly

undergo intersystem crossing to reach a lower-lying triplet state. Cryogenic rigid matrixes

do not prevent the intersystem crossing to the lower energy triplet states, thus singlet

carbenes and excited states can not be detected directly by this method.

1.4. Computational Chemistry in the 1970s

The year 1970 was proposed to be the starting date for the third age of quantum

chemistry .30 Computational chemistry established additional credibility in the 1970s upon

unraveling of the structure and energies of methylene, a molecule proposed as the

paradigm for computational quantum chemistry by Schaefer. 31 In the 1960s triplet

methylene was generally believed to be linear. 27 Even though the bent structure was

predicted by some calculations, 32 chemists around this time did not trust theoretical

results until they were confirmed by experimental data. However, in 1970 Bender and

Schaefer 27 challenged this idea by proposing a nonlinear geometry (135º) of triplet methylene on the basis of theoretical studies. This attracted much attention of chemists and quickly more experimental evidence accumulated that led to the reinterpretation of the original spectroscopic results 14 and confirmed the bent geometry. The currently

3 accepted experimental value is 133.84º.30 In 1970 chemists knew that methylene is a ground state triplet species, but disagreed on the singlet-triplet gap ( ∆EST ). Theorists

33 predicted a value of ∆EST = 11 ± 3 kcal/mol using various methods, but a simple direct

measurement in 1976 concluded this value to be 19.4 kcal/mol. 34 Finally, the computational prediction was proven to be correct by additional experimental studies.

Now the accepted value of ∆EST for methylene is 9 kcal/mol. This example, along with

others around this time, illustrated that theory can provide quantitatively accurate results,

especially for systems that are hard to study experimentally (reactive intermediates and

excited states). Since the 1980s, computational chemistry has become a mainstream part

of chemistry and has advanced rapidly over the years. Nowadays, quantum calculations

can reliably predict the singlet-triplet energy gaps for many carbenes, to forecast the UV

and IR spectra of carbenes and even predict carbene reaction rates.

1.5. Nanosecond Laser Flash Photolysis in the 1980s

Porter and Norrish were awarded the Nobel Prize in chemistry in 1967 for the

development of the flash photolysis method, 35 a powerful tool for the study of chemical reactions at ambient conditions. The potential of this method was later augmented by the advances in pulsed lasers and computers, and the time resolution of the experiment improved from the millisecond to nanosecond regime. Nanosecond laser flash photolysis

(ns-LFP) entered the toolbox of physical organic chemists in the late 1970s, 36 and is nowadays a common tool to study carbene chemistry in real-time.

The first absolute rate constant of a carbene reaction in solution was reported by

Closs and Rabinow in 1976. 37 In the 1980s there was widespread use of ns-LFP

4 technology in carbene studies. Many aryl- and diarylcarbenes, including fluorenylidene, 38-41 diphenylcarbene, 37,42,43 and 1- and 2-naphthylcarbene, 44,45 were detected in their triplet ground states, in solution, and these studies were interpreted with the aid of prior matrix isolation studies. However, most singlet arylcarbenes have lifetimes that are sub-nanosecond and can not be studied directly with ns-LFP techniques.

Arylhalocarbenes have singlet ground states with lifetimes on the order of a few microseconds 46 and were the first singlet carbenes detected by ns-LFP methodology. 47,48

Many alkyl, alkylhalo, dialkyl, and carbonyl carbenes, species which lack useful chromophores for direct UV detection, were indirectly studied by detecting their pyridinium ylides, an approach developed by Platz et al. in the late 1980s. 49 These studies

provided important insights into the unimolecular carbene-rearrangement. 50-55 However,

it is an indirect method and is often based on various assumptions. 56 Thus direct observation of singlet carbenes and the excited states of precursors is desired to increase our understanding and this has been achieved with ultrafast studies.

1.6. Ultrafast Spectroscopic Methods since the 1990s

The Nobel Prize in 1999 was awarded to Zewail for his studies of chemical reactions using femtosecond time-resolved spectroscopy. 57-59 In the last twenty years femtosecond time-resolved lasers have been used in the study of carbenes. Previously, singlet diphenylcarbene (DPC), 60 fluorenylidene (Fl), 61 and p-biphenylcarbene

(BpCH) 62,63 had been probed with ultrafast UV-Vis spectroscopy by the Kohler and Platz

research groups. In these studies, the excited states of diazo compounds were observed

and the mechanism of singlet carbene formation was explored. However, the excited

states of diazirines and diazo compounds were still not clearly understood.

5 1.7. Ultrafast and Theoretical Studies of Diazirine Excited states and Carbenes

In this dissertation, we will apply state-of-the-art ultrafast time-resolved UV-Vis and IR spectroscopies and ab initio quantum calculations to study the photochemistry of diazirine and diazo compounds and directly observe singlet carbenes.

Prior theoretical studies only dealt with simple or symmetric alkyl diazirines and

64-66 66 diazo compounds, for example, parent diazirine and (CH 2N2), and

67 dimethyldiazirine ((CH 3)2CN 2). Theory predicts that the excited state of diazirine

(CH 2N2) deactivates by passage through a conical intersection to form a diradical-like structure to form methylene. 66-68 However, the photochemistry of asymmetric

aryldiazirines, which have been studied recently in our group, 69-71 have not been previously studied. Chapter 2 presents a theoretical study of the excited states of

phenyldiazirine and phenyldiazomethane and the mechanism of their denitrogenation to

produce carbenes as well as their isomerization on the ground and excited state (S 1, S 2)

potential surfaces. These theoretical results are compared the experimental studies in

chapters 3 – 6 and deal with specific questions left unsolved by the experimental work.

In Chapter 3 we present the first direct observation of singlet phenylcarbene and

the measurement of its lifetime in solution using ultrafast time-resolved infrared (IR)

spectroscopy. 70 Some other singlet arylcarbenes were also detected by probing their C=C vibrational bands in various solvents. Photoisomerization is studied by monitoring diazo formation using its characteristic vibrational band. We conclude that both singlet carbene and diazo compound are born directly from the excited state (S 2) of the diazirine

precursor within the laser pulse (300 fs) and undergo vibrational cooling over < 100 ps

after the laser pulse.

6 In Chapter 4 we provide the first direct observation of the S 1 excited state of para -methoxy-3-phenyl-3-methyl diazirine ( p-CH 3OC 6H4CN 2CH 3) with both IR and

UV–vis detection techniques with 350 – 400 nm light.71 These experimental observations

are in excellent agreement with the prediction of calculations in Chapter 2 . It is

confirmed that the S1 state of diazirine decays into the diazo compound directly. We demonstrate that the S 2 excited state of aryldiazirine is populated with 270 nm light.

Denitrogenation into singlet arylcarbene and isomerization to diazo compound proceed

directly from the S 2 state and effectively compete with internal conversion to the S 1 state.

72 A Hammett study of diazirine S 1 excited states is presented in Chapter 5 . Six aryl substituted phenylchlorodiazirines were prepared and subject to ultrafast UV-Vis studies with light of 350 nm excitation wavelength. Excellent linear relationship in the

Hammett plot is obtained between the S 1 state lifetimes and the para - substituent

+ parameters σp in cyclohexane, acetonitrile, and chloroform. These studies permit a more precise representation of their excited states, as well as correlations of the excited state lifetimes with solvent polarity and the electronic properties of their aryl substituents, and are also consistent with computational studies (cf. Chapter 2 ) which suggest the S 1 excited states has lengthened C-N bonds, positive charge at the para and diazirine carbon

atoms, and negative charge at the nitrogen atoms.

We examined the effect of substitution of β-hydrogens on the S 1 state lifetimes (cf.

Chapter 6 ). The substitution effect and kinetic isotope effect in the nonpolar solvent cyclohexane are consistent with the Rearrangement In the Excited State (RIES) mechanism. The wavelength dependent effect on the excited states of p-biphenyldiazirine and p-biphenylmethyldiazirine was studied with 270 – 400 nm wavelength excitation

7 light. We provide direct evidence for the formation of the S 2 state with light of higher excitation energy and for the formation of the S 1 state with light of lower excitation

energy. Internal conversion from the S2 to the S 1 state was also detected.

In Chapter 7 we will present the first direct observation of a singlet vinylcarbene

and study its cyclization to a cyclopropene product in solution. 73 Vinylcarbene

PhCH=CHCCO 2CH 3 has a triplet ground state. In solution, both the cyclopropene product and the triplet vinylcarbene are directly detected with ultrafast UV-vis and IR spectroscopic methods. Studies of the kinetic isotope effect in CH3OD indicate that vinylcarbene also undergoes intramolecular reactions with alcohols. The intramolecular and intermolecular reactions of vinylcarbene are also studied with computational methods.

Calculations predict that both singlet and triplet vinylcarbene have highly delocalized structures over the C=C double bond.

The intramolecular C−H insertion reactions of carbenes are investigated in

Chapter 8 using ultrafast IR spectroscopic studies on N,N-diethyldiazoacetamide. 74 We

concluded that the excited state of diazoamide precursor undergoes direct intramolecular

C−H insertions in forming both β- and part of the γ -lactam products, as well as

denitrogenation to produce singlet carbene. This is the first direct ultrafast time resolved

evidence for the RIES mechanism. The relaxed singlet carbene decays by isomerizing

into γ-lactam in a concerted intramolecular C−H insertion in chloroform, and in CH 3OD this path is greatly suppressed by intermolecular reaction with solvent. DFT calculations predict a large solvent effect on the carbene carbonyl stretches, which explains the unusual observation that singlet carbene was not observed in chloroform, but observed in

CH 3OD.

8 1.8. References for Chapter 1

1. IUPAC Gold Book Website. http://goldbook.iupac.org/C00806.html (accessed February 4, 2010).

2. Kirmse, W. Carbene Chemistry, 2nd Ed. ; Academic Press: New York, 1971.

3. Dumas, J. B. Ann. Chim. Phys. 1835 , 58 , 28.

4. Regnault, H. V. Ann. Chim. Phys. 1839 , 71 , 427.

5. Butlerov, A. M. Ann. 1861 , 120 , 356.

6. Geuther, A. Ann. Chem. Pharm. 1862 , 123 , 121-122.

7. Platz, M. S. Observing Invisible Carbenes By Trapping Them with Pyridine. In Carbene Chemistry: From Fleeting Intermediates to Powerful Reagents ; Bertrand, G., Ed.; Marcel Dekker: New York, 2002; pp 27-56.

8. Buchner, E.; Feldmann, L. Ber. Dtsch. chem. Ges. 1903 , 36 , 3509-3517.

9. Staudinger, H.; Kupfer, O. Ber. Dtsch. Chem. Ges. 1912 , 45 , 501-509.

10. Doering, W. v. E.; Knox, L. H. J. Am. Chem. Soc. 1956 , 78 , 4947-4950.

11. Hine, J. J. Am. Chem. Soc. 1950 , 72 , 2438-2445.

12. Doering, W. v. E.; Hoffmann, A. K. J. Am. Chem. Soc. 1954 , 76 , 6162-6165.

13. Doering, W. v. E.; Buttery, R. G.; Laughlin, R. G.; Chaudhuri, N. J. Am. Chem. Soc. 1956 , 78 , 3224.

14. Herzberg, G.; Shoosmith, J. Nature 1959 , 183 , 1801-1802.

15. Nobel Foundation Website (1971). http://nobelprize.org/nobel_prizes/chemistry/laureates/1971/index.html (accessed February 4, 2010).

16. Woodworth, R. C.; Skell, P. S. J. Am. Chem. Soc. 1959 , 81 , 3383-3386.

17. Whittle, E.; Dows, D. A.; Pimentel, G. C. J. Chem. Phys. 1954 , 22 , 1943.

18. Becker, E. D.; Pimentel, G. C. J. Chem. Phys. 1956 , 25 , 224-228.

19. Norman, I.; Porter, G. Nature 1954 , 174 , 508-509.

20. Murray, R. W.; Trozzolo, A. M.; Wasserman, E.; Yager, W. A. J. Am. Chem. Soc. 1962 , 84 , 3213-3214.

9 21. Gibbons, W. A.; Trozzolo, A. M. J. Am. Chem. Soc. 1966 , 88 , 172-173.

22. Trozzolo, A. M.; Murray, R. W.; Wasserman, E. J. Am. Chem. Soc. 1962 , 84 , 4990-4991.

23. Trozzolo, A. M.; Gibbons, W. A. J Am Chem Soc 1967 , 89 , 239-243.

24. Ware, W. R.; Sullivan, P. J. J. Chem. Phys. 1968 , 49 , 1445-1446.

25. Closs, G. L.; Hutchisin, C. A., Jr.; Kohler, B. E. J. Chem. Phys. 1966 , 44 , 413-414.

26. Brandon, R. W.; Closs, G. L.; Hutchison, C. A., Jr. J. Chem. Phys. 1962 , 37 , 1878-1879.

27. Paulsen, S. R. Angew. Chem. 1960 , 72 , 781-782.

28. Schmitz, E.; Ohme, R. Chem. Ber. 1961 , 94 , 2166-2173.

29. Schmitz, E.; Ohme, R. Tetrahedron Lett. 1961 , 612-614.

30. Goddard, W. A., III. Science 1985 , 227 , 917-923.

31. Schaefer, H. F., III. Science 1986 , 231 , 1100-1107.

32. Foster, J. M.; Boys, S. F. Rev. Mod. Phys. 1960 , 32 , 305-307.

33. Harding, L. B.; Goddard, W. A., III. Chem. Phys. Lett. 1978 , 55 , 217-220.

34. Zittel, P. F.; Ellison, G. B.; Oneil, S. V.; Herbst, E.; Lineberger, W. C.; Reinhardt, W. P. J. Am. Chem. Soc. 1976 , 98 , 3731-3732.

35. Nobel Foundation Website (1967). http://nobelprize.org/nobel_prizes/chemistry/laureates/1967/ (accessed February 4, 2010).

36. Small, R. D., Jr.; Scaiano, J. C. J. Am. Chem. Soc. 1978 , 100 , 296-298.

37. Closs, G. L.; Rabinow, B. E. J. Am. Chem. Soc. 1976 , 98 , 8190-8198.

38. Brauer, B. E.; Grasse, P. B.; Kaufmann, K. J.; Schuster, G. B. J. Am. Chem. Soc. 1982 , 104 , 6814-6816.

39. Zupancic, J. J.; Schuster, G. B. J. Am. Chem. Soc. 1980 , 102 , 5958-5960.

40. Grasse, P. B.; Brauer, B. E.; Zupancic, J. J.; Kaufmann, K. J.; Schuster, G. B. J. Am. Chem. Soc. 1983 , 105 , 6833-6845.

41. Griller, D.; Montgomery, C. R.; Scaiano, J. C.; Platz, M. S.; Hadel, L. J. Am. Chem. Soc. 1982 , 104 , 6813-6814.

10 42. Hadel, L. M.; Platz, M. S.; Scaiano, J. C. J. Am. Chem. Soc. 1984 , 106 , 283-287.

43. Hadel, L. M.; Maloney, V. M.; Platz, M. S.; McGimpsey, W. G.; Scaiano, J. C. J. Phys. Chem. 1986 , 90 , 2488-2491.

44. Barcus, R. L.; Hadel, L. M.; Johnston, L. J.; Platz, M. S.; Savino, T. G.; Scaiano, J. C. J. Am. Chem. Soc. 1986 , 108 , 3928-3937.

45. Horn, K. A.; Chateauneuf, J. E. Tetrahedron 1985 , 41 , 1465-1470.

46. Moss, R. A.; Turro, N. J. In Kinetics and Spectroscopy of Carbenes and Biradicals ; Platz, M. S., Ed.; Plenum: New York, 1990; pp 213f.

47. Gould, I. R.; Turro, N. J.; Butcher, J., Jr.; Doubleday, C., Jr.; Hacker, N. P.; Lehr, G. F.; Moss, R. A.; Cox, D. P.; Guo, W.; et al. Tetrahedron 1985 , 41 , 1587-1600.

48. Turro, N. J.; Butcher, J. A.; Moss, R. A.; Guo, W.; Munjal, R. C.; Fedorynski, M. J. Am. Chem. Soc. 1980 , 102, 7576-7578.

49. Jackson, J. E.; Soundararajan, N.; Platz, M. S.; Liu, M. T. H. J. Am. Chem. Soc. 1988 , 110 , 5595-5596.

50. Modarelli, D. A.; Platz, M. S. J. Am. Chem. Soc. 1991 , 113 , 8985-8986.

51. Modarelli, D. A.; Morgan, S.; Platz, M. S. J. Am. Chem. Soc. 1992 , 114 , 7034- 7041.

52. Modarelli, D. A.; Platz, M. S.; Sheridan, R. S.; Ammann, J. R. J. Am. Chem. Soc. 1993 , 115 , 10440-10441.

53. Ford, F.; Yuzawa, T.; Platz, M. S.; Matzinger, S.; Fuelscher, M. J. Am. Chem. Soc. 1998 , 120 , 4430-4438.

54. Pezacki, J. P.; Couture, P.; Dunn, J. A.; Warkentin, J.; Wood, P. D.; Lusztyk, J.; Ford, F.; Platz, M. S. J. Org. Chem. 1999 , 64 , 4456-4464.

55. Liu, M. T. H.; Bonneau, R. J. Am. Chem. Soc. 1989 , 111 , 6873-6874.

56. Jackson, J. E.; Platz, M. S. Laser flash photolysis studies of ylide-forming reactions of carbenes. In Advances in Carbene Chemistry ; Brinker, U. H., Ed.; JAI Press: Greenwich CT, 1994; pp 89-160.

57. Zewail, A. H. Angew. Chem. Int. Ed. 2000 , 39 , 2586-2631.

58. Zewail, A. H. J. Phys. Chem. A 2000 , 104 , 5660-5694.

59. Nobel Foundation Website (1999). http://nobelprize.org/nobel_prizes/chemistry/laureates/1999/index.html (accessed February 4, 2010).

11 60. Peon, J.; Polshakov, D.; Kohler, B. J. Am. Chem. Soc. 2002 , 124 , 6428-6438.

61. Wang, J.; Kubicki, J.; Hilinski, E. F.; Mecklenburg, S. L.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2007 , 129 , 13683-13690.

62. Wang, J.; Burdzinski, G.; Gustafson, T. L.; Platz, M. S. J. Org. Chem. 2006 , 71 , 6221-6228.

63. Wang, J.; Burdzinski, G.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2007 , 129 , 2597-2606.

64. Mueller-Remmers, P. L.; Jug, K. J. Am. Chem. Soc. 1985 , 107 , 7275-7284.

65. Bigot, B.; Ponec, R.; Sevin, A.; Devaquet, A. J. Am. Chem. Soc. 1978 , 100 , 6575- 6580.

66. Yamamoto, N.; Bernardi, F.; Bottoni, A.; Olivucci, M.; Robb, M. A.; Wilsey, S. J. Am. Chem. Soc. 1994 , 116 , 2064-2074.

67. Bernardi, F.; Olivucci, M.; Robb, M. A.; Vreven, T.; Soto, J. J. Org. Chem. 2000 , 65 , 7847-7857.

68. Arenas, J. F.; Lopez-Tocon, I.; Otero, J. C.; Soto, J. J. Am. Chem. Soc. 2002 , 124 , 1728-1735.

69. Wang, J.; Burdzinski, G.; Kubicki, J.; Platz, M. S.; Moss, R. A.; Fu, X.; Piotrowiak, P.; Myahkostupov, M. J. Am. Chem. Soc. 2006 , 128 , 16446-16447.

70. Zhang, Y.; Burdzinski, G.; Kubicki, J.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 16134-16135.

71. Zhang, Y.; Burdzinski, G.; Kubicki, J.; Vyas, S.; Hadad, C. M.; Sliwa, M.; Poizat, O.; Buntinx, G.; Platz, M. S. J. Am. Chem. Soc. 2009 , 131 , 13784-13790.

72. Zhang, Y.; Wang, L.; Moss, R. A.; Platz, M. S. J. Am. Chem. Soc. 2009 , 131 , 16652-16653.

73. Zhang, Y.; Kubicki, J.; Platz, M. S. J. Am. Chem. Soc. 2009 , 131 , 13602-13603.

74. Zhang, Y.; Burdzinski, G.; Kubicki, J.; Platz, M. S. J. Am. Chem. Soc. 2009 , 131 , 9646-9647.

12 CHAPTER 2

THEORY OF PHENYLDIAZIRINE EXCITED STATES

2.1. Introduction

Diazirines, and their linear diazo isomers,1-3 are thermal and photochemical precursors of singlet carbenes.2-4 The mechanism of denitrogenation and formation of singlet carbenes from these precursors remain active areas of investigation. Recently, the photochemistry of aryl diazo compounds 5-9 and aryldiazirines 10,11 have been studied by

ultrafast time-resolved laser flash photolysis (LFP) techniques, and important insights

into the dynamics of precursor excited state decay and singlet arylcarbene formation have

been obtained. The excited states of p-biphenylyldiazomethane 5 and p-

biphenylyldiazoethane 6 were detected by ultrafast UV–vis spectroscopy, and it was found

that singlet arylcarbenes were produced directly from the decay of the diazo excited

states.5,6,12 However, the specific diazo excited state and its structure along with the decay

pathways on the electronically excited state potential energy surfaces were not previously

considered, and these issues will be addressed in this study.

Phenyldiazirine and the phenylchlorodiazirine analogue have been studied

previously by ultrafast infrared (IR) spectroscopy (270 nm excitation) techniques. 11 It

was reported that the excited states of phenyldiazirine undergo denitrogenation to

produce singlet phenylcarbene as well as photoisomerization to phenyldiazomethane

13 within 1 ps of the laser pulse. However, the specific excited states of phenyldiazirine responsible for these transformations were not characterized. A transient visible absorption band was also detected within the laser pulse of ultrafast LFP (350 nm) of arylhalodiazirines which was assigned to a ring-opened dipolar zwitterionic species.10

However, the nature of this zwitterionic species was not established, specifically whether it was an intermediate or an excited state of the diazirine, and also whether or not this species produces diazo compound or carbene. In a recent study of p-methoxy-3-phenyl-3-

13 methyldiazirine ( p-CH 3OC6H4CN 2CH 3) utilizing both UV–vis and IR spectroscopies, a transient species was observed and assigned to the C=C vibrational band of the S 1 excited state of the precursor as predicted by RI-CC2 calculations. Formation of the corresponding diazo compound and carbene was also confirmed from both the S 1 and S 2 excited states of the aryldiazirine. The fact that both carbene and diazo compounds can be formed from both the S 1 and S 2 states complicates the challenge of understanding of these

excited states.

Diazirines photoisomerize to diazo compounds ( Scheme 2.1), which can subsequently fragment to form carbenes. 14-16 Since only one C−N bond is broken during diazo formation (path a), and two C−N bonds are broken (path b) during carbene formation, simple intuition predicts that the formation of a diazo compound will be more facile than carbene formation. Indeed, our recent ultrafast LFP studies with aryl diazirines indicate that diazo formation is the dominant process in both the S 1 and S 2 states of aryldiazirines.13 This is not surprising, as Liu et al. reported evidence for the intermediacy of a diazo species in their study of the mechanism of the thermal decomposition of

alkylphenyldiazirines and. 4,17 However, the question of whether singlet carbenes are

14 directly formed from diazirine precursors in photochemical processes, or via the intermediacy of a diazo compounds remains to be answered. 13

Scheme 2.1. Two decomposition pathways of phenyldiazirine.

The direct formation of a carbene from a diazirine is also of inherent theoretical interest, because it can be viewed as the reverse of the classic carbene plus olefin cyclopropanation reaction. Based on the principle of microscopic reversibility, 18 the nitrogen extrusion process on the ground state surface can be viewed as the reverse of the reaction of a carbene and a nitrogen molecule ( Scheme 2.2).

Scheme 2.2. Two mechanisms proposed in the carbene addition to ethylene.

The reaction of singlet carbenes with olefins has received continuous interest. The streospecificity of cyclopropanation with singlet carbenes led to the postulation that simultaneous bond formation takes place between the carbon atoms of the singlet carbene and the olefin. 19 The mechanism has been subjected to numerous theoretical studies. 20-26

15 It is now generally accepted that a singlet carbene approaches ethylene in a non-least- motion fashion (B), rather than the symmetric, least-motion approach (A). 20 According to the Woodward-Hoffman rules, the cycloaddition of singlet carbene to ethylene is forbidden if C2V symmetry is preserved (A), and thus the asymmetric, non-least-motion

path (B) is symmetry allowed. 20 Further theoretical work 22 by Rondan, Houk, and Moss concluded that the cyclopropanation reaction begins with an electrophilic π approach (B), which is favored by overlap of the empty p-orbital of carbene with the π-orbital (HOMO) of ethylene. It is followed by nucleophilic σ approach (involving rotation of the carbene group), which overlaps the σ-orbital of the carbene with the π*-orbital (LUMO) of ethylene. The dihedral angle ( θ) between the carbene plane and the olefin plane is determined by the nucleophilicity of the singlet carbene ( Scheme 2.2). 22 Experimental evidence for the asynchronous formation of two C–C bonds have been obtained from

Singleton’s KIE studies. 1,27 The principle of microscopic reversibility 18 predicts that the

nitrogen extrusion process can be viewed as the reverse of the reaction of a carbene

attacking a nitrogen molecule; hence, the cleavage of the two C–N bonds may also be

non-synchronous in carbene formation from diazirine as the isomerization of

phenyldiazirine into phenyldiazomethane necessarily involves asymmetric cleavage of

one C–N bond. Thus, the partition between the two asynchronous processes of C–N

bonds cleavage in the carbene and diazo formation poses an interesting question.

The excited states of simple diazirines have been visited frequently with various

theoretical methods. 28-31 However, all of these investigations dealt with either the

smallest member of the diazirine functional group family, or symmetric dialkyldiazirines

(dimethyldiazirine, adamantyldiazirine), which may have different behavior than the

16 aryldiazirines. The photochemical reactions of asymmetric diazirines, especially aryldiazirines which have been studied recently in our group, 10,11,13 have not been

previously studied by theoretical means.

Herein, we present a theoretical study of the chemistry of phenyldiazirine and

phenyldiazomethane in their ground as well as in their excited states. Moreover, the

pathways for interconversion on both the ground and excited state surfaces will be

discussed. Finally, we will address different possible mechanisms involved in

photochemical routes. This theoretical work is in excellent agreement with the previously

reported experimental work and deals with specific questions left unsolved by the

experimental work.

2.2. Computational Methods

Ground state geometry optimizations were obtained with Becke’s three-parameter

hybrid exchange functional with the Lee-Yang-Parr correlation functional (B3LYP) 32-35 as implemented in Gaussian 03 36 and second-order coupled cluster methods with the resolution-of-the-identity approximation (RI-CC2), 37-39 as implemented in Turbomole

5.91 .40 Vibrational frequency analyses were performed to verify that the stationary points corresponded to energy minima (zero imaginary vibrational frequencies) or transition states (one imaginary vibrational frequency). The DFT vibrational frequencies were corrected with a scaling factor of 0.9614. 41 No scaling factor was used for the RI-CC2

frequencies. The electronic spectra were computed using the time-dependent B3LYP

(TD-B3LYP) methodology with Gaussian 03 at the B3LYP/6-311+G(d,p)//B3LYP/6-

31G(d) level. Vertical excitations were also calculated at the RI-CC2/TZVP level of

theory in Turbomole 5.91. The electronic difference density plots were generated by

17 calculating the electronic wave functions for the vertical excited states in Turbomole5.91 at the RI-CC2/TZVP level and visualized with UCSF Chimera.42,43 The optimized ground state geometry at the RI-CC2/TZVP level of theory was used as an initial input to optimize the respective geometries of the excited states. The stationary points obtained for the singlet excited states were confirmed to be minima by calculating the second derivatives numerically using the NumForce module in Turbomole 5.91. Each RI-

CC2/TZVP transition state located on the ground state surfaces of phenyldiazirine and phenyldiazomethane, was recomputed at the B3LYP/6-31+G(d) level of theory, and intrinsic reaction coordinate (IRC) 44,45 calculations were performed to confirm the reactant and product associated with these transition states. Natural Population Analysis

(NPA) 46 of charges was carried out on the optimized ground and excited state geometries.

Throughout the paper, bottom-of-the-well energies were used without zero-point vibrational energy (ZPE) correction. All calculations were performed at The Ohio

Supercomputer Center.

2.3. Results and Discussion

2.3.1. Phenyldiazirine

2.3.1.1. Ground state equilibrium geometry, vertical excitations, and electronic difference density plots of phenyldiazirine

The singlet ground state of phenyldiazirine was optimized at the B3LYP/6-

31+G(d) and RI-CC2/TZVP levels of theory and confirmed to be a minimum by frequency analysis. At both levels of theory, the optimized geometries have CS symmetry, and optimizations without symmetry constraints generated the same structure. Calculated vertical excitations with the TD-B3LYP and RI-CC2 methods on the optimized singlet

18 ground state geometries are summarized in Table 2.1 along with the oscillator strengths and orbitals involved in the transition.

TD-B3LYP/6-311+G(d,p)// RI-CC2/ TZVP B3LYP/6-31+G(d) Character Energy/eV Oscillator Character Energy/eV Oscillator State (% contribution) (nm) strength (% contribution) (nm) strength

S1 31 →32 (95%) 3.19 (388) 0.0064 31 →32 (52%) 3.73 (332) 0.0049 29 →32 (5%) 29 →32 (24%)

S2 30 →32 (97%) 4.51 (281) 0.0052 30 →33 (27%) 5.06 (245) 0.0098 31 → 33 (3%) 31 →34 (27%) 31 →33 (17%) 30 →32 (13%)

S3 31 →34 (35%) 5.05 (245) 0.0063 30 →32 (45%) 5.95 (208) 0.0475 29 → 32 (23%) 31 →33 (23%) 30 → 33 (23%) 31 →34 (11%) 31 →33 (13%) 30 →33 (10%)

Table 2.1. Vertical excitation energies, oscillator strengths, and the dominant occupied to virtual orbital configurations (>10%) contributing to the three lowest energy singlet excitations of phenyldiazirine calculated at the TD-B3LYP/6-311+G(d,p)//B3LYP/6- 31+G(d) and RI-CC2/TZVP levels of theory.

Figure 2.1. Steady state UV–vis absorption spectrum of phenyldiazirine in pentane.

19 TD-B3LYP calculations ( Table 2.1) predict that the three lowest singlet excited

states are well separated and have similar oscillator strengths. The results are validated by

the good correspondence between the calculated and measured UV–vis transition

energies ( Figure 2.1). Therefore, we propose that 350 nm excitation pumps

phenyldiazirine to the S 1 state, and 270 nm light promotes phenyldiazirine to the S 2 state;

moreover, the S 3 state (245 nm) is not accessible by 270 – 400 nm excitation. The results of the RI-CC2 calculations are similar, except that the vertical excitation energies are overestimated, as reported earlier. 47,48

34 33 32

31 30 29

Figure 2.2. Graphic representation of representative molecular orbitals of phenyldiazirine contributing to the three lowest energy singlet excitations calculated at the B3LYP/6- 311+G(d,p) level of theory.

The contour surfaces and ordering of molecular orbitals obtained by the B3LYP

calculations are presented in Figure 2.2. The orbitals involved in the three lowest energy

20 excitations are similar in both methods. Calculations predict that the S 1 state is largely

due to a HOMO →LUMO transition for which the HOMO (31) is localized in the σ-bond of the two C–N linkages, and the LUMO (32) is the π*-orbital of the N=N bond. Similar

31 results were also predicted for dialkyldiazirines. The S 2 state is characterized as a π

(phenyl) → π* (phenyl) transition with little contribution from the diazirine subunit, while the S 3 state involves different π-orbitals of the aromatic unit but has considerable

contribution from the diazirine subunit as well. However, because these transitions have

multiconfigurational character, visualization of the orbitals does not necessarily provide

insight into the potential photochemistry, and one way to represent the electronic

redistribution is to consider the differences in the total electron densities of the ground

and the excited states. Difference electron density plots for the S 1, S 2 and S 3 excited states are shown in Figure 2.3.

S1 state S2 state S3 state (isocontour value ±0.005) (isocontour value ±0.001) (isocontour value ±0.003)

Figure 2.3. Excited state difference density plots for the S 1, S 2 and S 3 states relative to the ground state electron density for phenyldiazirine, as calculated at the RI-CC2/TZVP level of theory. A red surface surrounds areas where electron density is depleted after vertical excitation from S 0; a green surface surrounds areas where electron density is accumulated in the excited state.

In these plots, a red contour represents a depletion of electron density from the ground state, and a green contour depicts an accumulation of electron density in the

21 excited state. It is clear that the S 1 state is a σ→π* state localized on the diazirine group, and there is no significant change in electron density in the phenyl group. Loss of electron density in the σC–N bond and the accumulated electron density in the π*N=N bond

of the diazirine group suggests cleavage of the C–N bonds and formation of diazo

compound or fragmentation to molecular nitrogen and phenylcarbene from the S 1 state.

These results are consistent with previous theoretical studies of parent diazirine and

dimethyldiazirine that concluded that the first singlet excited state results from an

30,31,49 electronic σ→π* transition. The S 2 and S3 states are predicted to be largely π→π*

in nature and localized primarily on the benzene ring, with some contribution from the

diazirine group. Predicting the possible photochemistry on S 2 and S 3 excited states from

the difference density plots alone is not straightforward to.

2.3.1.2. Excited State Equilibrium Geometries of Phenyldiazirine

The equilibrium geometries of the S 1, S 2, and S 3 excited states of phenyldiazirine

were optimized at the RI-CC2/TZVP level of theory and confirmed to be minima by

evaluation of the Hessian matrix for each state; thus, each stationary point on the excited

state potential energy surface yielded only real vibrational frequencies ( Figure 2.4).

In the S 1 state, the two C–N bonds of the diazirine group are remarkably

elongated by 0.083 and 0.088 Å relative to the ground state, and the N=N bond length is

also elongated by 0.04 Å. The lengthening of these bonds is consistent with depletion of

electron density in the bonding orbital ( σC–N) or accumulation of electron density in the anti-bonding orbital ( π*N=N) as depicted in the density plots ( Figure 2.3). A significant

31 C–N bond lengthening was also predicted in the S 1 state of dimethyldiazirine. Of particular interest is that the S 1 state shows a noticeable difference (0.005 Å) between the

22 two C–N bond lengths and has C1 symmetry, while the predicted ground state has CS symmetry.

S0 S1

S2 S3

Figure 2.4. Optimized geometries for the ground and electronic excited states of phenyldiazirine at the RI-CC2/TZVP level of theory. Distances are shown in angstroms.

Prior studies at the CASSCF level predicted that the S 1 states of parent diazirine

30,31 (CH 2N2) and dimethyldiazirine have equivalent C–N bonds. Geometry optimizations of the S1 state of phenyldiazirine with CS geometry yielded a stationary point that, upon

computation of the Hessian, had an imaginary vibrational frequency, which corresponds

to the asymmetric stretch of the two C–N bonds. Hence, we conclude that phenyldiazirine

is slightly deformed in the S 1 state. The C–C bond lengths of the phenyl moiety indicate a

50 quinoidal structure of the S 1 state. The bond pattern is similar to that of benzyl cation.

23 Consistent with this speculation, charge analysis indicates that a charge of +0.1 e develops at C 12 (the diazirine ring carbon, Table 2.2), and negative charges mainly resides on ( 0.11 e and 0.13 e), relative to the S 0 ground state.

atom S0 S1 S2 S3 1C 0 -0.02 -0.04 -0.03 2C 0 0.04 0.03 0.01 3C 0 -0.05 -0.06 -0.13 4C 0 0.05 -0.01 0.03 5C 0 -0.01 -0.02 -0.04 6C 0 0.05 0.00 0.01 7H 0 0.01 0.01 0.00 8H 0 0.02 0.01 0.01 9H 0 0.02 0.00 0.01 10H 0 0.01 0.01 0.01 11H 0 0.01 0.01 0.01 12C 0 0.10 0.05 0.16 13H 0 0.01 0.01 0.02 14N 0 -0.11 0.00 0.01 15N 0 -0.13 -0.01 -0.09

Table 2.2. The charge analysis of S 1, S 2 and S 3 states relative to S 0 state of phenyldiazirine optimized at the RI-CC2/TZVP level of theory. Charges on the ground state were subtracted to show the changes in the excited states. Positive values indicate that the atom is more positively charged than in the ground state, while negative charges indicate that atom is more negatively charged than in the ground state.

In the S 2 state, the two C–N bonds lengths are predicted to be the same length

(CS) and are lengthened by 0.023 Å compared to the ground state. In addition, all C=C

24 bonds of the phenyl ring are lengthened, consistent with a π→π* state localized on the phenyl group.

Surprisingly, the optimized S 3 state has a ring-opened structure ( Figure 2.4). One of the C–N bonds is increased by 0.326 Å compared to S 0 and the other C–N bond is

almost unchanged. Charge analysis predicts that the S3 state has the largest amount of charge separation ( Table 2.2), with a positive charge of +0.16 e on the diazirine carbon

(C 12 ) and a negative charge of –0.09 e on nitrogen (N 15 ) relative to the ground state. This state has all of the characteristics predicted for the zwitterionic structure proposed for the polar intermediate detected in the ultrafast UV–vis studies of arylhalodiazirines. 10

However, we believe that the S 3 state is not the zwitterion observed in the experiments

because calculations predict that the S0→S3 transition is at 245 nm ( Table 2.1), and can

not be pumped with 350 nm excitation. Furthermore, it is unreasonable to posit that the

S3 state has a relatively long lifetime (5–800 ps) as we would expect rapid internal

13 conversion (less than 1 ps) to the lower energy excited states (S 2 and S 1).

A summary of the geometric parameters, ground and excited states energies and dipole moments of the optimized structures are provided in Table 2.3. Inspection of the calculated dipole moments reveals that the S 1 state has an exceedingly large dipole moment (4.75 D), even though the S 3 state is the most ionic species. The predicted large

12 dipole moment of the S 1 state is consistent with the experimental observation of a red-

shift on the absorption spectra by substitution and the longer lifetimes of the transient

species in the polar solvent. Therefore, we believe that ultrafast LFP (350 nm) of

phenyldiazirine creates the S 1 state. Consistent with this assignment, TD-B3LYP

calculations predict the S 1 state is populated with 350 nm excitation ( Table 2.1).

25 Furthermore, structural sampling on the ground state and excited state potential surfaces did not locate any other zwitterionic structures, except the inaccessible S3 state.

Dipole States Energies C-N N=N N -C -N C -C -H Moment 15 12 14 3 12 13

S0 1.82 0 1.493/1.493 1.258 49.81 120.02

S1 4.75 80.8 1.581/1.576 1.298 48.56 122.10

S2 1.62 111.9 1.516/1.516 1.253 48.80 120.93

S3 1.20 117.5 1.819/1.490 1.233 42.3 124.74

Table 2.3. Dipole moments (debye), relative energies (kcal/mol), and some geometric parameters of S 0, S 1, S 2 and S 3 states of phenyldiazirine optimized at the RI-CC2/TZVP level of theory.

States Vibrational frequency (intensity)

S0 1575 (28.6), 1616 (66.2), 1649 (3.9), 1720 (1.0), 1756 (2.9)

S1 1513 (37.7), 1586 (0.7), 1607 (764.5)

S2 1519 (54.2), 1561 (18.8), 1677 (38.8)

S3 1507 (600.0), 1664 (75.9), 1854 (21.4)

Table 2.4. Vibrational frequencies (cm -1) and their respective intensities (km/mol) of the ground and excited states of phenyldiazirine computed at the RI-CC2/TZVP level of theory.

As shown in Table 2.4, a strong C=C vibrational mode at 1607 cm -1 is predicted

for the S 1 state of phenyldiazirine, while the S 2 excited state and the S 0 ground state are not expected to have a significant band in the region of 1500 – 3000 cm -1. Thus, on the basis of these theoretical calculations, we predict that the 1607 cm -1 band can be used as

26 an IR marker for the S 1 state. We have shown that for p-methoxy-3-phenyl-3-

methyldiazirine, the same vibrational band is predicted and was directly detected by

13 ultrafast infrared spectroscopy. This confirms the proposition that the S 1 state of diazirine was the carrier of the transient observed by both ultrafast UV–vis and IR spectroscopic techniques, and validates the computational methods employed in this study.

2.3.2. Phenyldiazomethane

2.3.2.1. Ground state equilibrium geometry, vertical excitations, and electronic difference density plots of phenyldiazomethane

The ground state geometry of phenyldiazomethane was optimized without constraint at the B3LYP/6-31+G(d) and RI-CC2/TZVP levels of theory and confirmed to be a minimum. Calculated vertical excitations and details are summarized in Table 2.5.

TD-B3LYP calculations predict that the transition energy to the S 1 state of

phenyldiazomethane (506 nm) is smaller than that predicted for phenyldiazirine (388

nm), and with very small oscillator strength ( vide infra , cf Figure 2.8). The predicted S 2

(283 nm) and S 3 states (268 nm) are very close in energy, but the oscillator strength of the

S3 state is more than an order of magnitude larger than that of the S 2 state. This result

suggests that 270 – 310 nm excitations probably will populate the S 3 state directly. The

RI-CC2 method predicts similar results except for an over estimation of the excitation

energies.

27

TD-B3LYP/6-311+G(d,p) RI-CC2/ TZVP //B3LYP/6-31+G(d) Character Energy/eV Oscillator Character Energy/eV Oscillator States (%contribution) (nm) strength (%contribution) (nm) strength 31 → 32 (93%) 2.45 (506) 0.0000 31 → 33 (81%) 2.86 (434) 0.0000 S 1 29 → 33 (12%) 31 → 34 (70%) 4.38 (283) 0.0353 31 → 34 (59%) 4.68 (265) 0.0298

S2 31 → 33 (14%) 30 → 32 (17%) 30 → 33 (10%) 31 → 32 (16%) 31 → 33 (61%) 4.63 (268) 0.4514 31 → 32 (76%) 5.04 (246) 0.5432 S 3 31 → 34 (14%) 31 → 34 (12%)

Table 2.5. Vertical excitation energies, oscillator strengths, and the dominant occupied to virtual orbital configurations (>10%) contributing to the three lowest energy singlet excitations of phenyldiazomethane calculated at the TD-B3LYP/6-311+G(d,p)// B3LYP/6-31+G(d) and RI-CC2/TZVP levels of theory.

34 33 32

31 30 29

Figure 2.5. Graphic representation of representative molecular orbitals of phenyldiazomethane contributing to the three lowest energy singlet excitations calculated at the B3LYP/6-311+G(d,p) level of theory.

28 The contour surfaces and the ordering of the molecular orbitals obtained by the

TD-B3LYP method are shown in Figure 2.5. The orbitals involved in the three lowest energy excitations are found to be similar in both methods. For example, the TD-B3LYP and the RI-CC2 methods predict that the S 1 state is the transition from the out-of-plane

bonding orbital ( πN=N ) to the in-plane anti-bonding orbital (π*N=N ). Both the S 2 state and

S3 states have considerable contribution from the π-orbital of the phenyl group.

S1 state S2 state S3 state (isocontour value ±0.005) (isocontour value ±0.001) (isocontour value ±0.001)

Figure 2.6. Excited state difference density plots for the S 1, S 2 and S 3 states relative to the ground state electron density for phenyldiazomethane, as calculated at the RI- CC2/TZVP level of theory. A red surface surrounds areas where electron density is depleted after vertical excitation from S 0; a green surface surrounds areas where electron density is accumulated in the excited state.

Electronic difference density plots between the excited states and the ground state are shown in Figure 2.6. These plots reveal that the S 1 state of phenyldiazomethane

involves promoting an electron from the πC=N orbital to the in-plane π*N=N orbital. This is consistent with the cleavage of C–N bond and formation of phenylcarbene and molecular nitrogen. Therefore, we conclude that the S 1 state of phenyldiazomethane is a dissociative

state. Later, we will show that promotion of phenyldiazirine to its S 1 state is also

predicted to produce the S 1 state of phenyldiazomethane, which then fragments to form

29 carbene and decays to the ground state of phenyldiazomethane. In the S 2 and S 3 states, a

considerable depletion of the electron density from the πC=N orbital is also observed,

albeit the π-orbitals in the phenyl group are also significantly involved. So, we can

predict that there is a channel for formation of phenylcarbene on the S 2 and S 3 states.

2.3.2.2. Excited State Equilibrium Geometries of Phenyldiazomethane

The equilibrium geometries of the S 1, S 2, and S 3 excited states of

phenyldiazomethane were optimized at the RI-CC2/TZVP level of theory and confirmed

to be minima by computing the Hessian ( Figure 2.7)

Geometry optimization of the S 1 state of phenyldiazomethane indicates that the

molecule maintains CS symmetry. The most significant change is that the C–N–N bond is bent (130˚) in S 1 relative to being almost linear (177˚) in the ground state (S 0). Hence two

possible isomers can exist on the S 1 state surface, cis and trans . The trans isomer is found to be more stable by 1.3 kcal/mol than the cis isomer, probably due to steric effects. The decreased bond angle (C–N–N) is consistent with the πC=N →π*N=N transition being localized on the diazo group. This change in bond angle is similar to a rehybridization effect due to an increase in p character. A decreased bond angle has also been shown for

51 47 the excited states of CO 2, formaldehyde, alkyl and aryl .

In the S 1 state, the C–N bond is elongated by 0.147 Å. The increased C–N bond length and decreased bond angle indicates that the S 1 state is a dissociative state and will probably have a low barrier to denitrogenation relative to the ground state. The N=N bond length is elongated by 0.034 Å, consistent with the population of the π*N=N orbital

in the excited state. The C ipso –Cbenzylic bond length is shortened by 0.045 Å, suggesting

delocalization of electron density from the aromatic group to the benzylic carbon in the

30 S1 state. Bond length alteration in the phenyl group is small, but nonetheless resembles a quinoidal structure.

S0

S1-trans S1-cis

S2 S3

Figure 2.7. Optimized geometries of the ground and excited states of phenyldiazomethane at the RI-CC2/TZVP level of theory. Bond lengths are in Angstroms, and bond angles are in degrees.

The geometry of the S2 state also remains CS symmetric, and the C=N=N diazo

group remains approximately linear, similar to the S 0 state. There are some slight

increases in the bond lengths of the C=N=N group, even though this effect is less

significant than in the S 1 state. However, all of the C–C bond lengths in the phenyl group

are elongated by about 0.017 to 0.038 Å. This is consistent with a π→π* state localized

on the phenyl group.

31 In the S 3 state, the geometry is significantly different from the ground and other singlet excited states as this stationary point has only C1 symmetry. The diazo group is

bent out of the phenyl plane with an angle of 160˚, and the dihedral angle C–C–N=N is

close to 90˚. This structure is of particular interest because it resembles the ring-opened

structure of phenyldiazirine (derived from its S 3 state, cf. Figure 2.4). Hence we predict that the S 3 state of phenyldiazomethane is possibly connected to the diazirine manifold, most likely, the S 3 state. The C=N bond is increased by 0.053 Å, and the C=C bond lengths in the phenyl group are increased to a different extent, in agreement with the S 3 state being a combination of π→π* states on both the phenyl and diazo moieties.

Figure 2.8.The composite energy diagram of the vertical and adiabatic excited states of phenyldiazirine (left) and phenyldiazomethane (right), as calculated at the RI-CC2/TZVP level of theory. The energies are in kcal/mol and are relative to the S 0 ground state of phenyldiazomethane.

The relative energy level diagram of the ground and excited states of phenyldiazomethane and phenyldiazirine is summarized in Figure 2.8. The S 3 state of

32 phenyldiazomethane is lower in energy than the S 3 state of phenyldiazirine. Therefore we

propose that the S 2 state of phenyldiazirine can easily isomerize into the S 3 state of the diazo compound. The S 1 state of the diazo compound is much lower in energy than the S 1 state of phenyldiazirine. Thus the photochemical isomerization of phenyldiazirine to phenyldiazomethane is an exothermic process on the S 1 state surface.

2.3.3. Potential Energy Surfaces Scan

2.3.3.1 Ground State and S 1 Excited State Potential Energy Surfaces Scan of Phenyldiazirine

Since both the formation of carbene and diazo compound from phenyldiazirine

involves the cleavage of the C–N bonds of the diazirine group, we performed a potential

energy surface scan for the ground (S0, Figure 2.9, left panel) and S 1 ( Figure 2.9, right

panel) excited states of phenyldiazirine along the C–N bond. As the C 12 –N15 bond was

stretched, the other internal coordinates were fully optimized at the RI-CC2/TZVP level

of theory.

As shown in Figure 2.9, the potential energies increase as the C 12 –N15 bonds

deviate from the equilibrium geometry (dashed line) on both the S 0 and S 1 surfaces.

Interestingly, the changes on the second C–N bond (C 12 –N14 ) length are quite different for these two states. The second C–N bond (C 12 –N14 ) elongates gradually on the S 0 surface (left panel), but it shortens substantially on the S 1 surface, as the C12 –N15 bond is

stretched (right panel, Figure 2.9). The symmetric and asynchronous lengthening of both

C-N bonds on the ground state is indicative of phenylcarbene formation. The asymmetric lengthening of one C–N bond and shortening of the other on the S 1 surface (right panel)

of phenyldiazirine is consistent with phenyldiazomethane formation via

photoisomerization. Meanwhile, the shortening of the N=N bond length and growth of

33 the C–N–N bond angle are also consistent with fragmentation on the ground state (left panel) and isomerization on the S 1 state (right panel). Indeed, the optimization of the transition states ( TS1 and TS2 ) on reaction paths predicts that TS1 leads to the ground state phenylcarbene on the ground state surface (left panel), and TS2 leads to the S 1 state

of phenyldiazomethane on the S 1 state of phenyldiazirine (right panel). Thus we conclude that phenylcarbene formation on the ground state and diazo formation (S 1) on the S 1 excited state of phenyldiazirine are facile based on these results.

34

Figure 2.9. The relaxed optimized geometries of phenyldiazirine after extension of the C12 –N15 bond at the RI-CC2/TZVP level of theory, starting from the ground (S 0, left panel) and S 1 excited (right panel) state optimized geometries. The transition structures of TS1 (ground state) and TS2 (S 1 excited state) obtained are shown on top. Energies are relative to the ground state equilibrium geometry of phenyldiazirine in kcal/mol. Bond distances are in Angstroms. Bond angles are in degrees.

The transition state ( TS1 , Figure 2.9) leading to phenylcarbene and N 2 formation was found with an activation energy of 28.7 kcal/mol on the ground state surface for

phenyldiazirine. Frey and Stevens investigated the thermal unimolecular decomposition

of dimethyldiazirine, and it was found to be first-order reaction, giving propene and

35 nitrogen, 52 consistent with computational results that decomposition of diazirine into carbene on the ground state surface is facile. Liu et al. measured the activation energy for the denitrogenation of arylalkyldiazirines by thermolysis to be 30 kcal/mol, 17 which is in

very good agreement with these gas-phase predictions. An activation energy of 31.0

kcal/mol for dimethyldiazirine was located by Soto et al. at the CASSCF/6-31G(d) level

of theory in the formation of dimethylcarbene on the ground state surface.31

Figure 2.10. The relaxed optimization of phenyldiazirine by fixing one C–N bond (black curve) and two C–N bonds (red curve) on the ground (S 0), S 1, and S 2 excited states potential energy surfaces at the RI-CC2/TZVP level of theory. Energies are relative to the ground state equilibrium geometry of phenyldiazirine in kcal/mol. Bond distances are in Angstroms.

The mechanism of carbene formation is of particular interest. As shown in Figure

2.9, one C–N bond progresses faster than the other one during the scan. In addition, the

C–N–N bond angle grows from 65 ° to 100 ° and the dihedral angle C 4–C3–C12 –N14 gradually changes from –29 ° to 37 °, indicating that the perpendicular bisector of N=N subunit moves out and leans toward the plane of phenylcarbene subunit (cf. TS1 of

36 Figure 2.9). As discussed in the introduction, carbene formation from diazirine can be viewed as the reverse of asynchronous cyclopropanation. The results are consistent with

an asynchronous pathway. In addition, the transition state structure TS1 obtained is also similar to the transition state for carbene addition to an alkene. 22 The synchronous

elongation of the two C–N bonds was also performed and indeed, the potential energy

curve of this process is higher than that of the asynchronous pathway ( Figure 2.10 ) for

phenyldiazirine S 0, S 1, and S 3 states. Therefore, we conclude that the asynchronous carbene formation is preferred for the S 0 ground state of phenyldiazirine.

The above calculations do not predict the formation of diazo compound from

ground state diazirine. Since diazo formation also involves the cleavage of one of the two

C–N bonds, the mechanism of diazo and carbene formation via two different

asynchronous processes is an intriguing possibility. A relaxed scan was performed by

increasing the C–N–N bond angle on the ground surface to explore this process.

However, the attempt to find the transition state for this process failed. Based on these,

we concluded that singlet phenylcarbene formation is probably preferred over diazo

formation on the ground state surface of phenyldiazirine.

Contrary to the ground state data, the stretching of one C–N bond on the S 1 state surface is accompanied by the contraction of the second C–N bond. These geometric

changes are indicative of ring-opening of the diazirine and subsequent formation of the

diazo compound. In accord with this result, compressing the C 12 –N15 bond also leads to lengthening of the C 12 –N14 bond. Re-optimization of the end points in both directions without geometry constraint of the S 1 state of phenyldiazirine leads to the S 1 state of phenyldiazomethane ( trans conformer). The transition state ( TS2 , Figure 2.9) with an

37 activation energy barrier of only 2.5 kcal/mol was found for this transformation. This result suggests that the S 1 state of phenyldiazirine has a very low energy barrier for isomerization to phenyldiazomethane. If a normal pre-exponential factor of 10 –13 is assumed, the Arrhenius equation predicts that the S1 state of phenyldiazirine has a lifetime of 6.6 ps at 300 K, which is in excellent agreement with the 5 ps (in acetonitrile)

lifetime determined by ultrafast UV–vis spectroscopic methods. 53 Previous CASSCF calculations on dimethyldiazirine predict a transition state of 1.0 kcal/mol from the S 1

31 30 state minimum. Similar results were also predicted with 3 H-diazirine (CH 2N2).

However, our calculations do not predict the direct formation of phenylcarbene

from the S 1 state of phenyldiazirine. A potential energy scan of the C–N–N bond angle failed to locate a transition state for the carbene formation. Previously we have reported the direct observation of phenylcarbene generated upon 270 nm excitation of phenyldiazirine. 11,13 To reconcile the calculations and experimental findings we propose

that the photochemical singlet carbene formation upon excitation of phenyldiazirine

proceeds via the intermediacy of the S 1 state of phenyldiazomethane ( vide infra ).

2.3.3.2 Ground State and S 1 Excited State Potential Energy Surface Scans of Phenyldiazomethane

The elongation of the C–N bond in the ground state of phenyldiazomethane leads

to an increase in potential energy and eventually to the formation of phenylcarbene and

molecular nitrogen ( Figure 2.11 ). The decreased N=N bond length and C–N–N bond

angle is consistent with this prediction. A transition state ( TS3 ) is located 33.9 kcal/mol

above the reactant for the S 0 state, and corresponds to formation of phenylcarbene from

the ground state of phenyldiazomethane.

38

Figure 2.11. The relaxed optimized geometries of phenyldiazomethane after extension of the C 12 –N15 bond at the RI-CC2/TZVP level of theory, starting from the S 0 ground state and leading to the TS3 transition state for extrusion of molecular nitrogen. Energies are relative to the ground state equilibrium geometry in kcal/mol. Bond distances are in Angstroms. Bond angles are in degrees. Atom labeling is provided in TS3 structure to the right.

The thermal denitrogenation of the ground state diazo compound does not follow the simple pathway of linearly stretching the C–N bond under CS symmetry. Rather, C 12 –

N14 shortens as C 12 –N15 is stretched and proceeds to the transition state. The C 12 –N14 –N15 group bends from linear to ~ 148 °, and the dihedral angle C 2–C3–C12 –N14 increases from

0° to ~ 90 ° in transition state TS3 , before it forms singlet carbene and molecular nitrogen

(Figure 2.11 ). This results indicate that the N=N subunit moves from its linear geometry

in the phenyldiazomethane plane toward a near perpendicular bisector position of the

phenylcarbene unit during denitrogenation. This process is certainly reminiscent of the

39 denitrogenation of phenyldiazirine, and both processes of carbene formation are non- least-motion fashion.

Figure 2.12. The contour plot of the potential energy surface of phenyldiazirine along the two C–N bond coordinates, as computed at the B3LYP/6-31+G(d) level of theory.

A complete potential energy surface scan on the ground state of phenyldiazirine

and phenyldiazomethane was further explored with Gaussian03 . As shown in Figure

2.12 , the valley at the lower right corner, with an equal value of C–N bonds at 1.49 Å,

corresponds to phenyldiazirine; while the other two valleys with C–N bond lengths of 1.3

and 2.4 Å are phenyldiazomethane. The shallow valley at the top left corner consists of

two very elongated C–N bonds corresponding to phenylcarbene and molecular nitrogen.

The reaction paths (TS1 and TS3 ) are indicated as dashed lines. It is clear that the

formation of phenylcarbene from phenyldiazirine does not follow the symmetric,

synchronous pathway ( Figure 2.12 , dashed lines along TS1 ), because the diagonal line

40 represents a high energy pathway. Instead, the reaction path diverges from the diagonal, and one C–N bond elongates faster than the other, and then the second C–N bond increases faster and reaches the transition state (TS1 ). From phenyldiazomethane, the

reaction path follows a 45º slope at an early stage, and curves from this slope at longer

C–N bond lengths, indicating that elongation of the C–N bond is linear at early stage, and

non-symmetric at longer bond lengths before it reaches the transition state ( TS3). This

plot also indicates that phenyldiazirine is surrounded with high-energy barriers, and the

isomerization from diazirine to diazo is not the preferred pathway. This is consistent with

the predictions of the RI-CC2 calculations.

Methods Phenyldiazirine TS1 TS3 Singlet Phenylcarbene + N 2 RI-CC2/TZVP 13.3 42.0 34.0 29.9 B3LYP/6-31G(d) 12.6 44.1 38.4 36.4 B3LYP/6-311+G(d,p) 15.2 43.7 35.8 31.5 MP2/6-311+G(df,p) 6.9 43.3 35.9 33.6

Table 2.6. Calculated transition states energies and related species with various methods. Energies are in kcal/mol relative to phenyldiazomethane. ZPE is not included. Singlet point energy with the MP2 method were calculated with the geometry optimized at the B3LYP/6-311+G(d,p) level of theory. TS1 is the transition state of carbene formation from phenyldiazirine, and TS3 is the transition state of carbene formation from phenyldiazomethane.

The transition states of carbene formation from phenyldiazirine ( TS1 ) and

phenyldiazomethane ( TS3 ) were further computed using B3LYP and MP2 methods and were confirmed with IRC calculations. The relative energies from different levels of theory were presented in Table 2.6 and the structures in Figure 2.13 .

41 TS1 TS3 TS1 42.0 TS3 29.9 34.0

13.3

0

Figure 2.13. The thermal transition states for the formation of phenylcarbene from phenyldiazirine (TS1) and phenyldiazomethane (TS3) calculated at the RI-CC2/TZVP level of theory on the ground state surfaces. Bond lengths are in Angstroms.

The elongation of the C–N bond of the S 1 state of phenyldiazomethane leads to

the formation of phenylcarbene and molecular nitrogen. The transition state was not

found but the energy barrier is estimated to be ~2.7 kcal/mol. For the trans -S1 state, the

barrier is even smaller, as little as 0.3 kcal/mol before the S 1 state fragments. As shown above, isomerization into phenyldiazomethane is facile on the S 1 surface of

phenyldiazirine, with a low energy barrier (2.5 kcal/mol, TS2 , Figure 2.9). The energy

difference between the S 1 surface of phenyldiazirine and S 1 surface of

phenyldiazomethane is predicted to be about 50 kcal/mol ( Figure 2.8). Hence, we posit that the S 1 state of phenyldiazomethane is born with excess vibrational energy from phenyldiazirine, and as a result fragmentation and carbene formation will be rapid. An

42 economical interpretation of phenylcarbene formation from phenyldiazirine 11,13 is passage through the TS2 (2.5 kcal/mol) to form the S 1 state of phenyldiazomethane, which consequently forms phenylcarbene. This is consistent with the prediction that singlet carbene was formed from the excited state of phenyldiazirine via the intermediacy of phenyldiazomethane.

2.4. Conclusions

Phenyldiazirine and phenyldiazomethane were studied at the TD-B3LYP/6-

31+G(d) and RI-CC2/TZVP levels of theory, and the three lowest singlet excited states of both compounds were optimized using the RI-CC2/TZVP method. The calculations predict that the S 1 state of phenyldiazirine is a σ→π* state, with a quinoidal structure for the phenyl group, and the C–N bonds of the diazirine group are slightly deformed from the CS symmetry of the ground state geometry. However, the S2 and S 3 states are

predicted to be π→π* states on the phenyl moiety. Among the examined electronic states,

the S 1 state was predicted to have an exceedingly large dipole moment and a strong aromatic C=C vibrational mode around ~1600 cm -1. The calculations are consistent with an assignment of the S 1 state of the diazirine to the polar intermediate recently observed

by ultrafast time-resolved UV–vis and IR spectroscopic studies of arylhalo- and

arylalkyldiazirines. 10,11 The ring-opened zwitterionic structure of the transient species, as

10 proposed in the previous studies, forms on the S 3 surface which is inaccessible with 270 nm pump excitation. The excited states of phenyldiazomethane were also studied and the mechanism of interconversion between phenyldiazirine and phenyldiazomethane were discussed. The calculations predict that the chemistry of the ground and the S 1 excited

43 states of phenyldiazirine is very different. On the ground state surface, the formation of phenylcarbene is favored and on the S 1 excited state, the formation of phenyldiazomethane in the first excited state is greatly favored. The S 1 state of phenyldiazirine can isomerize into the S 1 state of phenyldiazomethane over a very low

energy barrier. The S 1 state of phenyldiazomethane can rapidly fragment into phenylcarbene.

2.5. References for Chapter 2

1. Moss, R. A.; Platz, M. S.; Jones, M., Jr. Reactive Intermediate Chemistry ; Wiley: New York, 2004.

2. Liu, M. T. H. Chemistry of Diazirines ; CRC Press, 1987.

3. Moss, R. A. Acc. Chem. Res. 2006 , 39 , 267-272.

4. Jennings, B. M.; Liu, M. T. H. J. Am. Chem. Soc. 1976 , 98 , 6416-6417.

5. Wang, J.; Burdzinski, G.; Gustafson, T. L.; Platz, M. S. J. Org. Chem. 2006 , 71 , 6221-6228.

6. Wang, J.; Burdzinski, G.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2007 , 129 , 2597-2606.

7. Wang, J.; Kubicki, J.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 2304-2313.

8. Wang, J.; Burdzinski, G.; Kubicki, J.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 11195-11209.

9. Peon, J.; Polshakov, D.; Kohler, B. J. Am. Chem. Soc. 2002 , 124 , 6428-6438.

10. Wang, J.; Burdzinski, G.; Kubicki, J.; Platz, M. S.; Moss, R. A.; Fu, X.; Piotrowiak, P.; Myahkostupov, M. J. Am. Chem. Soc. 2006 , 128 , 16446-16447.

11. Zhang, Y.; Burdzinski, G.; Kubicki, J.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 16134-16135.

12. Wang, J.; Kubicki, J.; Hilinski, E. F.; Mecklenburg, S. L.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2007 , 129 , 13683-13690.

44 13. Zhang, Y.; Burdzinski, G.; Kubicki, J.; Vyas, S.; Hadad, C. M.; Sliwa, M.; Poizat, O.; Buntinx, G.; Platz, M. S. J. Am. Chem. Soc. 2009 , 131 , 13784-13790.

14. Nigam, M.; Platz, M. S.; Showalter, B. M.; Toscano, J. P.; Johnson, R.; Abbot, S. C.; Kirchhoff, M. M. J. Am. Chem. Soc. 1998 , 120 , 8055-8059.

15. Bonneau, R.; Liu, M. T. H. J. Am. Chem. Soc. 1996 , 118 , 7229-7230.

16. Bonneau, R.; Liu, M. T. H.; Kim, K. C.; Goodman, J. L. J. Am. Chem. Soc. 1996 , 118 , 3829-3837.

17. Liu, M. T. H.; Jenning, B. M. Can. J. Chem. 1977 , 55 , 6.

18. Chandrasekhar, S. Res. Chem. Intermed. 1992 , 17 , 173-209.

19. Skell, P. S.; Garner, A. Y. J. Am. Chem. Soc. 1956 , 78 , 3409-3411.

20. Hoffmann, R. J. Am. Chem. Soc. 1968 , 90 , 1475-1485.

21. Zurawski, B.; Kutzelnigg, W. J. Am. Chem. Soc. 1978 , 100 , 2654-2659.

22. Rondan, N. G.; Houk, K. N.; Moss, R. A. J. Am. Chem. Soc. 1980 , 102 , 1770- 1776.

23. Houk, K. N.; Rondan, N. G. J. Am. Chem. Soc. 1984 , 106 , 4293-4294.

24. Blake, J. F.; Wierschke, S. G.; Jorgensen, W. L. J. Am. Chem. Soc. 1989 , 111 , 1919-1920.

25. Keating, A. E.; Garcia-Garibay, M. A.; Houk, K. N. J. Am. Chem. Soc. 1997 , 119 , 10805-10809.

26. Bernardi, F.; Bottoni, A.; Canepa, C.; Olivucci, M.; Robb, M. A.; Tonachini, G. J. Org. Chem. 1997 , 62 , 2018-2025.

27. Keating, A. E.; Merrigan, S. R.; Singleton, D. A.; Houk, K. N. J. Am. Chem. Soc. 1999 , 121 , 3933-3938.

28. Mueller-Remmers, P. L.; Jug, K. J. Am. Chem. Soc. 1985 , 107 , 7275-7284.

29. Bigot, B.; Ponec, R.; Sevin, A.; Devaquet, A. J. Am. Chem. Soc. 1978 , 100 , 6575- 6580.

30. Yamamoto, N.; Bernardi, F.; Bottoni, A.; Olivucci, M.; Robb, M. A.; Wilsey, S. J. Am. Chem. Soc. 1994 , 116 , 2064-2074.

31. Bernardi, F.; Olivucci, M.; Robb, M. A.; Vreven, T.; Soto, J. J. Org. Chem. 2000 , 65 , 7847-7857.

45 32. Becke, A. D. J. Chem. Phys. 1993 , 98 , 5648-5652.

33. Becke, A. D. J. Chem. Phys. 1992 , 96 , 2155-2160.

34. Becke, A. D. J. Chem. Phys. 1992 , 97 , 9173-9177.

35. Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988 , 37 , 785-789.

36. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A. J.; T. Vreven; K. N. Kudin; J. C. Burant; J. M. Millam; S. S. Iyengar; J. Tomasi; V. Barone; B. Mennucci; M. Cossi; G. Scalmani; N. Rega, G. A. P., H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al- Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez, and J. A. Pople. Gaussian 03, Revision E.01 ed.; Gaussian, Inc.: Wallingford CT, 2004.

37. Hattig, C. J. Chem. Phys. 2003 , 118 , 7751-7761.

38. Haettig, C.; Koehn, A. J. Chem. Phys. 2002 , 117 , 6939-6951.

39. Kohn, A.; Hattig, C. J. Chem. Phys. 2003 , 119 , 5021-5036.

40. Ahlrichs, R.; Baer, M.; Haeser, M.; Horn, H.; Koelmel, C. Chem. Phys. Lett. 1989 , 162 , 165-169.

41. Scott, A. P.; Radom, L. J. Phys. Chem. 1996 , 100 , 16502-16513.

42. Molecular graphics images were produced using the UCSF Chimera package from the Resource for Biocomputing, Visualization, and Informatics at the University of California, San Francisco (supported by NIH P41 RR-01081)

43. Pettersen, E. F.; Goddard, T. D.; Huang, C. C.; Couch, G. S.; Greenblatt, D. M.; Meng, E. C.; Ferrin, T. E. J. Comput. Chem. 2004 , 25 , 1605-1612.

44. Gonzalez, C.; Schlegel, H. B. J. Chem. Phys. 1989 , 90 , 2154-2161.

45. Gonzalez, C.; Schlegel, H. B. J. Phys. Chem. 1990 , 94 , 5523-5527.

46. Reed, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys. 1985 , 83 , 735-746.

46 47. Wang, J.; Kubicki, J.; Burdzinski, G.; Hackett, J. C.; Gustafson, T. L.; Hadad, C. M.; Platz, M. S. J. Org. Chem. 2007 , 72 , 7581-7586.

48. Burdzinski, G.; Hackett, J. C.; Wang, J.; Gustafson, T. L.; Hadad, C. M.; Platz, M. S. J. Am. Chem. Soc. 2006 , 128 , 13402-13411.

49. Arenas, J. F.; Lopez-Tocon, I.; Otero, J. C.; Soto, J. J. Am. Chem. Soc. 2002 , 124 , 1728-1735.

50. Laube, T.; Olah, G. A.; Bau, R. J. Am. Chem. Soc. 1997 , 119 , 3087-3092.

51. Turro, N. J. Modern Molecular Photochemistry ; University Press: Menlo Park, CA, 1978.

52. Frey, H. M.; Stevens, I. D. R. J. Chem. Soc. 1962 , 3865-3867.

53. Zhang, Y.; Platz, M. S. Org. Lett. 2010 , in preparation .

47 CHAPTER 3

DIRECT OBSERVATION OF CARBENE AND DIAZO FORMATION FROM ARYLDIAZIRINES BY ULTRAFAST INFRARED SPECTROSCOPY

This chapter is reproduced with permission from J. Am. Chem. Soc. 2008 , 130 , 16134. Copyright 2008 American Chemical Society.

3.1. Introduction

Diazirines are a class of organic molecules containing a three-membered ring,

which consists of a tetravalent carbon atom and an N=N double bond (Scheme 3.1).

These compounds have received increasing interest since their discovery in the early

1 2,3 1960s by Paulsen, Schmitz and Ohme. Aryl- and alkyl- substituted diazirines (R 1 and

R2 in Scheme 3.1) can be conveniently synthesized from their respective or ketones. 2,4 The pioneering work of Graham improved the synthesis of parent diazirine

5-7 8 (CH 2N2) and a host of halodiazirines, which in turn can be converted into a wide variety of diazirines through exchange reactions. 9 Diazirines are resistant to bases and

strong acids, 3 in sharply contrast to their linear isomer diazo compounds. Unlike their highly strained cyclopropene analogs, diazirines are remarkably stable in solution at room temperature. Nevertheless, diazirines produce carbenes in high yields with either heat or light ( Scheme 3.1). Therefore, diazirines are excellent carbene sources and have played an important role in carbene chemistry. 9,10

48 h or

N N N h or h or N + N R1 R2 2 R1 R2 R1 R2 Diazirine Diazo compound Carbene Scheme 3.1. The classical chemical reactions of diazirines.

Early studies of Frey and Stevens reported the formation of methylene (CH 2) by

11 the photolysis of parent diazirine (R 1 = R 2 = H, Scheme 3.1) in the gas phase.

Phenylchlorocarbene was produced and detected spectroscopically upon irradiation of phenylchlorodiazirine in a cryogenic matrix. 12 Carbenes can be captured by alkenes in solution of diazirines upon photolysis. 13,14 Pyridine ylide formation has been used to

study “invisible” carbenes produced from diazirines in laser flash photolysis (LFP)

experiments. 15 Amrich and Bell identified diazomethane as a product in the photolysis of parent diazirine. 16 The formation of diazo compounds in the photolysis of diazirines was reported in cryogenic matrixes 17 and in solution at room temperature 18-21 with various spectroscopic techniques, and was predicted by theoretical studies. 22 Therefore the

formation of both carbenes and diazo compounds from the photolysis of diazirines is

unquestioned. However, it is still unclear whether carbenes are formed directly from

diazirines, or via the intermediacy of diazo compound produced in the photolysis of

diazirines. 16,23

Ultrafast time-resolved (fs, ps) laser flash photolysis (LFP) studies with UV–vis and infrared (IR) detection are needed to understand early events in the photochemistry of carbene precursors. We have recently reported the use of ultrafast UV–vis spectroscopic studies of arylhalodiazirines 24 and IR studies of carbonyl diazo compounds

and their related carbenes. 25 Herein, the application of ultrafast IR spectroscopy to the

49 study of the photochemistry of simple aryldiazirines and the observation of singlet carbenes and diazo compound are described.

3.2. Ultrafast Spectroscopic Results

3.2.1. 3-Chloro-3-phenyldiazirine

The one-pot halogenation of amidines, known as the Graham’s reaction, 8,9 provides a convenient synthesis of a variety of arylhalodiazirines (ArCN 2X).

Arylhalocarbenes (ArCX) were central to the study of bimolecular carbene reactions of

arylhalodiazirines using laser flash photolysis (LFP) studies. 26 Thus, it was

straightforward to apply ultrafast IR techniques to this class of compounds.

50

Figure 3.1. Transient IR spectra produced by photolysis of chlorophenyldiazirine ( λex = 270 nm) in chloroform. (a) The diazo band in the region 2060 – 1940 cm-1 with time delays of 3 – 236 ps. (b) The carbene band in the region 1630 – 1550 cm-1 with time delays of 4 – 164 ps.

Ultrafast LFP (270 nm) of chlorophenyldiazirine (PhCN 2Cl) in chloroform

produces the transient IR spectra shown in Figure 3.1. The negative band detected at

1568 cm -1 of the transient difference spectra is attributed to depletion of ground state

chlorophenyldiazirine ( Figure 3.1b) upon excitation. Two broad bands are observed within a few picoseconds of the laser pulse in the spectral range 2100 – 1550 cm -1

(Figure 3.1). These bands narrow and shift to higher wavenumber over 100 ps, then center at 2040 and 1583 cm -1 and maintain their intensities over the detection time

51 window (3 ns). The blue shift and band narrowing are characteristic of vibrational cooling (VC) of intermediates born with excessive vibrational energy. 27

The carrier of the transient band at 2040 cm -1 can be readily assigned to chlorophenyldiazomethane, the diazo isomer of chlorophenyldiazirine produced upon photolysis. 19 The 1583 cm -1 band is attributed to a C=C vibrational mode of singlet chlorophenylcarbene ( 1PhCCl), formed by denitrogenation of phenylchlorodiazirine.

Chlorophenylcarbene has been detected in its singlet ground state by irradiation of

chlorophenyldiazirine isolated in an argon matrix at 10 K by Sheridan et al.12 The

transient band absorbing at 1583 cm -1 is consistent with the intense IR band detected at

1590 cm -1 of 1PhCCl in a matrix. 12 Another IR band of 1PhCCl at 1225 cm -1 in a matrix 12 is not observed in this study due to the limitation of our spectral window (generally 2800

– 1400 cm -1). Toscano et al. detected 1PhCCl in solution at 1586 cm -1 using ns time-

resolved IR spectroscopy, 28 in good agreement with this study. Singlet chlorophenylcarbene has a lifetime of a few microseconds ( µs) at room temperature, 26,28 hence accounting for the constant intensity over 100 – 3000 ps in this study.

The mechanism of carbene and diazo compound formation from a diazirine precursor upon excitation was not understood well in prior ns time-resolved LFP studies, 28 and will be studied with the ultrafast time-resolved (fs) IR spectroscopy.

However, the narrowing and blue-shifting of transient IR bands within ~ 100 ps post laser pulse are due to VC and do not afford direct kinetic information on the ps time scale. The formation kinetics of transient IR bands can be obtained from the integration of band intensity, based on the assumption that VC process has no impact on the integrated spectra. This analysis has been performed by other groups, 29 and in our recent studies. 30,31

52

Figure 3.2. Transient IR spectra produced by photolysis of chlorophenyldiazirine (λex = 270 nm) in chloroform. The integration of the diazo band (2060 – 1940 cm -1) versus time delay (note the time delays in logarithmic scale).

The diazo band integration shown in Figure 3.2 clearly indicates that the

integrated diazo band in the 2060 – 1940 cm -1 spectral region has constant intensity over the 1 – 1000 ps time scale. Therefore we conclude that chlorophenyldiazomethane is formed instantaneously, within the laser pulse (< 1 ps). Based on calculations described in Chapter 2 we propose that the diazo compound is formed directly from the diazirine

1 electronic S 2 excited state. However, the growth of carbene PhCCl can not be accurately

determined because the C=C vibrational band of hot carbene 1PhCCl # severely overlaps

bleaching centered at 1568 cm -1 and strong background absorption of solvent below 1550 cm -1.

3.2.2. Phenyldiazirine

Ultrafast LFP ( λex = 270 nm) of phenyldiazirine in acetonitrile produces the transient IR spectra shown in Figure 3.3. A broad band in the region 2100 – 1900 cm -1 was formed within a few picoseconds of the laser pulse. It narrows and shifts to the blue at 2064 cm -1 over ~ 100 ps. Then this band maintains the same intensity within the

53 detection time window (3 ns) and is confidently assigned to the diazo band of phenyldiazomethane which is born with excessive vibrational energy and undergoes VC within 100 ps of the laser pulse.

Figure 3.3. Transient IR spectra produced by photolysis of phenyldiazirine ( λex = 270 nm) in chloroform. The diazo band in the spectral region of 2100 – 2035 cm -1 in a time window of 2 – 2200 ps.

Another positive transient IR band is observed in the spectral region of 1640 –

-1 1550 cm by ultrafast LFP ( λex = 270 nm) of phenyldiazirine in chloroform ( Figure 3.4).

Initially a broad band ~ 1560 cm -1 is formed within 1 ps of the laser pulse; then it grows

and narrows with a blue shift and reaches its maximum intensity at 50 ps post the laser

pulse.

54

Figure 3.4. Transient IR spectra produced by ultrafast LFP of phenyldiazirine ( λex = 270 nm) in chloroform. (a) The formation of singlet carbene 1PhCH in a time window of 1 – 50 ps. (b) The decay of singlet carbene 1PhCH band at 1582 cm -1 in a time window of 50 – 1200 ps.

Figure 3.5. The decay trace of phenylcarbene 1PhCH band monitored at 1582 cm -1 produced by ultrafast LFP of phenyldiazirine ( λex = 270 nm) in chloroform and was fit to an exponential function.

55 The transient IR band peaking at 1582 cm -1 decays exponentially with a lifetime

of 414 ± 74 ps in chloroform ( Figure 3.4b and Figure 3.5). Three negative bands at

1625, 1608 and 1580 cm -1 were also observed immediately after the laser pulse due to depletion of the ground state diazirine ( Figure 3.4). The carrier of the transient absorption band at 1582 cm -1 can be assigned to singlet phenylcarbene ( 1PhCH). DFT

1 -1 calculations predict that singlet PhCH has an intense νC=C vibration at 1571 cm ( Table

3.2), in good agreement with experiment. This is also consistent with singlet chlorophenylcarbene ( 1PhCCl) observed at 1583 cm -1 ( Figure 3.1b) and singlet

1 -1 25 NpCCO 2CH 3 which has a νC=C vibrational band at 1584 cm . This is the first direct spectroscopic observation of singlet 1PhCH and it also provides the first direct measurement of its lifetime.

The identification of triplet phenylcarbene ( 3PhCH) has been firmly established as

the ground state since the classic experiments of Wasserman and coworkers using

electron spin resonance (ESR) spectroscopy in low-temperature matrices. 32 Subsequent

independent experimental 33 and computational studies 34-36 have confirmed this finding and found that many aryl- or alkyl- substituted carbenes are triplet ground state species.

Thus, the decay of singlet 1PhCH in solution will lead to the lower energy triplet 3PhCH

via intersystem crossing (ISC). B3LYP/6-31+G(d) calculations predict that triplet 3PhCH

-1 3 has its νC=C vibration at 1543 cm . However, triplet carbene PhCH was not observed in

this experiment, presumably due to its low intensity (calculated to be 0.2, Table 3.2), compared with that of singlet 1PhCH (99.3, Table 3.2). Singlet carbene can also

deactivate through reacting with chloroform solvent, forming a chlorobenzyl radical

(PhCHCl, Scheme 3.2). 37 However, the radical PhCHCl was not observed in this

56 -1 experiment. DFT calculations predict the radical PhCHCl νC=C at 1550 cm with very

low intensity (calculated to be 0.2).

Scheme 3.2. The possible reaction pathways from singlet phenylcarbene from prior studies.

57

E = -270.2337355 Hartree C 2.48331000 -0.17343000 0.00095500 C 1.04965400 -0.04155100 -0.00056800 C 0.35063500 1.20022800 -0.00063000 C 0.27770600 -1.23568400 -0.00055300 C -1.03849800 1.24341700 0.00010500 H 0.92708900 2.12264600 -0.00112100 C -1.11314100 -1.19610300 -0.00003600 H 0.81836000 -2.17776500 -0.00080000 C -1.76628000 0.04339100 0.00050400 H -1.56290800 2.19535100 0.00016300 H -1.69277100 -2.11512400 0.00040100 H -2.85333700 0.07745800 0.00136000 H 2.90326200 0.85582600 0.00132900 Frequency IR intensity Frequency IR intensity Frequency IR intensity (cm -1) (km/mol) (cm -1) (km/mol) (cm -1) (km/mol) 177 5.3 937 2.6 1315 15.3 290 35.2 964 0.1 1425 26.4 376 10.7 974 0.8 1458 13.1 444 0.2 984 0.0 1545 1.9 519 2.2 1004 2.5 1571 99.3 522 43.1 1040 18.1 2818 121.5 601 0.4 1108 7.1 3060 0.1 659 45.8 1147 3.1 3068 3.9 749 51.4 1155 18.9 3078 17.3 802 0.2 1239 144.2 3085 18.3 827 0.3 1294 6.1 3096 7.2

Table 3.1. Structure of singlet phenylcarbene ( 1PhCH) optimized at the B3LYP/6- 31+G(d) level of theory (frequencies scaled by 0.9614) and cartesian coordinates.

58

E = -270.2414824 Hartree

C 2.444786000 -0.142368000 -0.000076000 C 1.052813000 -0.073194000 0.000040000 C 0.375315000 1.189754000 0.000015000 C 0.247609000 -1.254114000 -0.000023000 C -1.011622000 1.251232000 -0.000044000 H 0.963326000 2.103765000 -0.000002000 C -1.137795000 -1.171183000 0.000030000 H 0.741592000 -2.221606000 -0.000052000 C -1.780301000 0.076943000 0.000023000 H -1.503792000 2.220903000 -0.000043000 H -1.728955000 -2.083674000 -0.000002000 H -2.865283000 0.133949000 -0.000001000 H 3.248281000 0.584234000 0.000313000

Frequency IR intensity Frequency IR intensity Frequency IR intensity (cm -1) (km/mol) (cm -1) (km/mol) (cm -1) (km/mol) 194 1.5 853 15.4 1306 0.1 331 3.4 862 3.0 1409 3.2 388 0.4 932 0.0 1441 5.5 437 36.3 950 0.0 1518 1.2 486 1.6 951 0.5 1543 0.2 513 0.1 1000 3.5 3056 2.9 599 0.1 1067 3.8 3063 2.9 653 35.3 1139 0.1 3074 17.4 722 58.3 1148 0.5 3080 30.1 795 1.2 1255 0.9 3087 12.8 796 0.0 1267 0.3 3125 1.3

Table 3.2. Structure of triplet phenylcarbene ( 3PhCH) optimized at the B3LYP/6- 31+G(d) level of theory (frequencies scaled by 0.9614) and cartesian coordinates.

It is well known that singlet phenylcarbene undergoes intramolecular ring expansion reactions at elevated temperatures ( Scheme 3.2). 38 In the 1980s the Chapman group reported the detection of a species with IR bands at 1824 and 1816 cm -1, which was assigned to 1,2,4,6-cycloheptatetraene, following irradiation of singlet phenylcarbene isolated in a cryogenic matrix. 39,40 These assignment was substantiated later with more

59 comprehensive experimental 40-42 and computational work. 40 However, we did not observe the formation of this cycloheptatetraene species by ultrafast IR spectroscopy in this work.

This was not terribly surprising because this ring expansion reaction has only been reported in cryogenic matrices 39,40 or during thermolysis in the gas phase. 38 Recently,

Warmuth and Marvel achieved the direct detection of cyclcoheptatetraene incarcerated in a hemicarcerand from photochemical ring-expansion of isolated phenylcarbene at room temperature, albeit in very low yied. 43

3.2.3. Phenylmethyldiazirine

Phenylmethyldiazirine (PhCN 2CH 3) was also studied by ultrafast LFP ( λex = 270 nm) and transient IR absorption methods. The spectra are shown in Figure 3.6. The diazo compound produced by photoisomerization again undergoes vibrational cooling (VC) as evident by the narrowing and blue-shifting IR bands detected after laser excitation

(Figure 3.6a). The diazo IR band maintains its maximum intensity at 2040 cm -1 over 100

– 3000 ps post the laser pulse, similar to the observation for phenyldiazirine with 270 nm

excitation.

60

Figure 3.6. The transient spectra were generated by ultrafast LFP (270 nm) of phenylmethyldiazirine in chloroform. (a) The diazo band detected in the 2060 – 1930 cm - 1 spectral widow with time delays 1 – 100 ps. (b) The carbene band detected in the 1640 – 1550 cm -1 spectral widow with time delays of 2 – 30 ps

1 However, the intensity of the singlet phenylmethylcarbene ( PhCCH 3) C=C

vibrational band is greatly reduced relative to parent phenylcarbene (Figure 3.6b). As a

1 result, its lifetime can not be accurately determined. We posit that singlet PhCCH 3 might

1 undergo 1,2-H shift and consequently the concentration of PhCCH 3 is lower than that of

parent phenylcarbene, which lacks this intramolecular decay reaction. Alternatively, the

previously proposed RIES mechanism in which the olefin is formed directly from a

1 diazirine excited state constitutes another possibility for the low PhCCH 3 concentration. 44 However, both propositions are based on the assumption that the

1 1 absorption coefficients of PhCCH 3 and PhCH are similar. This assumption seems reasonable because the same νC=C vibrational mode of the phenyl group is observed for both carbenes. However, calculations predict that the absorption of this vibrational mode

1 1 -1 in PhCCH 3 is almost threes times weaker than that of PhCH (33.6 at 1569 cm , Table

61 1 3.3). Therefore, we conclude that the weak absorption observed for PhCCH 3 is at least partially due to its intrinsic small absorption coefficient ( vide infra ).

E = -309.5601762 Hartree C 1.883736000 -0.699975000 -0.260720000 C 0.506001000 -0.283547000 -0.119628000 C 0.047763000 1.062319000 -0.113550000 C -0.468748000 -1.312201000 -0.040030000 C -1.310563000 1.354779000 -0.038754000 H 0.761697000 1.875635000 -0.203812000 C -1.821358000 -1.017306000 0.106893000 H -0.118912000 -2.339399000 -0.088218000 C -2.244179000 0.317503000 0.092369000 H -1.649297000 2.387653000 -0.066651000 H -2.550550000 -1.818345000 0.195732000 H -3.303472000 0.551318000 0.168195000 C 2.926304000 0.249760000 0.191144000 H 3.285916000 -0.201000000 1.135587000 H 3.804854000 0.218667000 -0.466098000 H 2.656028000 1.297483000 0.418923000 Frequency IR intensity Frequency IR intensity Frequency IR intensity (cm -1) (km/mol) (cm -1) (km/mol) (cm -1) (km/mol) 75 12.6 920 3.3 1388 6.4 131 9.0 954 0.9 1417 10.8 198 14.8 968 0.4 1444 14.8 242 2.9 975 0.2 1464 1.0 399 4.7 1003 8.9 1543 0.3 409 3.5 1022 8.7 1569 33.6 459 2.4 1042 5.7 2834 74.0 536 2.1 1074 3.6 2889 42.2 600 0.4 1145 1.1 2960 10.4 662 32.7 1160 21.2 3057 1.1 736 40.3 1238 154.8 3068 9.9 760 14.5 1272 0.6 3078 20.1 823 5.7 1310 8.5 3085 18.4 825 2.7 1318 42.2 3092 10.0

1 Table 3.3. Structure of phenylmethylcarbene ( PhCCH3) optimized at the B3LYP/6- 31+G(d) level of theory (frequencies scaled by 0.9614) and cartesian coordinates.

62 3.2.4. p-Biphenyldiazirine

p-Biphenyldiazirine ( p-BpCN 2H) was studied by ultrafast LFP ( λex = 270 nm) and transient IR spectra demonstrating p-biphenyldiazomethane formation are shown in

Figure 3.7a. Again, the vibrationally excited diazo band is formed within 1 ps of the

laser pulse and exhibits a broad band peaking at 2040 cm -1. Then, the hot diazo

compound undergoes vibrational cooling as the broad band narrows and shifts to higher

wave numbers over ~ 100 ps; finally it centers at 2086 cm -1 and maintains its intensity

over 3 ns. Integration of the diazo band intensities in the 2110 – 2020 cm -1 spectral region

demonstrates, however, that the concentration of the diazo compound is constant within 1

– 3000 ps ( Figure 3.7b). The data indicate that the diazo compound is formed within 1 ps of the laser pulse, consistent with the results of diazo formation from chlorophenyldiazirine upon photolysis.

Figure 3.7. Ultrafast LFP (λex = 270 nm) of p-biphenyldiazirine in acetonitrile. (a) Transient spectra of diazo compound in a time window of 2 – 180 ps. (b) The integration of the diazo band intensities versus time delays.

63 1 Singlet p-piphenylcarbene ( BpCH) was also observed by ultrafast LFP ( λex = 270

nm) of p-biphenyldiazirine in chloroform ( Figure 3.8). Hot singlet carbene 1BpCH undergoes vibrational cooling within 60 ps post laser pulse; then centers at 1570 cm -1 and

decays to baseline over 1.5 ns. The intensity of 1BpCH is stronger than that of 1PhCH

1 carbene. Calculations predict that p-biphenylcarbene ( BpCH) has a stronger νC=C vibrational band than does 1PhCH, in excellent agreement with the experimental data

(Table 3.4).

Figure 3.8. Transient spectra were generated by ultrafast LFP (270 nm) of p- biphenyldiazirine in chloroform. (a) Transient IR spectra showing the growth of 1BpCH carbene in a time window of 4 – 65 ps. (b) Transient IR spectra showing the decay of 1BpCCH carbene in a time window of 65 – 1500 ps.

64

E = -501.2770211 Hartree C -1.112870000 1.149826000 0.363568000 C -2.497262000 1.142711000 0.366265000 C -3.237500000 -0.014925000 -0.003403000 C -2.489645000 -1.165098000 -0.371561000 C -1.103842000 -1.165772000 -0.368243000 C -0.388394000 -0.006812000 -0.002130000 H -0.572051000 2.038120000 0.676729000 H -3.042416000 2.035584000 0.663741000 H -3.053499000 -2.045545000 -0.665167000 H -0.558840000 -2.051345000 -0.681009000 C -4.673299000 -0.114568000 -0.041965000 H -5.066097000 0.870188000 0.305101000 C 1.092774000 -0.000596000 -0.000425000 C 1.810119000 1.149869000 -0.374079000 C 1.817850000 -1.145681000 0.374827000 C 3.203342000 1.153328000 -0.376547000 H 1.271549000 2.036759000 -0.694964000 C 3.211032000 -1.138775000 0.380062000 H 1.284993000 -2.036500000 0.694241000 C 3.909398000 0.009885000 0.002502000 H 3.738315000 2.048772000 -0.680837000 H 3.751971000 -2.030292000 0.685304000 H 4.995862000 0.013903000 0.003628000 Frequency IR intensity Frequency IR intensity Frequency IR intensity (cm -1) (km/mol) (cm -1) (km/mol) (cm -1) (km/mol) 67 2.2 821 9.0 1291 3.9 70 0.8 829 2.8 1307 3.0 111 1.5 837 24.5 1321 3.2 174 7.1 902 1.0 1395 31.0 273 25.0 937 1.1 1436 2.3 284 1.4 950 0.0 1467 3.7 289 8.3 963 0.0 1491 11.0 346 4.1 969 0.1 1516 9.6 401 4.3 978 3.3 1573 0.8 402 1.7 983 2.8 1580 344.3 489 3.9 999 0.9 1597 4.3 506 31.2 1027 0.7 2787 190.9 547 2.6 1064 19.6 3062 8.9 555 2.2 1074 2.5 3063 1.9 607 0.2 1121 1.4 3069 4.5 624 0.6 1149 0.0 3076 11.9 684 19.6 1161 41.8 3079 5.6 700 5.3 1173 24.3 3082 4.3 718 17.3 1244 209.2 3085 33.0 755 35.4 1263 40.2 3091 19.3 814 1.0 1266 1.9 3096 6.3

Table 3.4. Structure of singlet p-biphenylcarbene ( 1BpCH) optimized at the B3LYP/6- 31+G(d) level of theory (frequencies scaled by 0.9614) and cartesian coordinates.

65 The kinetic trace probed at 1570 cm -1 shows that 1BpCH decays with a lifetime of

599 ps in chloroform ( Figure 3.9a), which is slightly longer than the lifetime of 1PhCH

(τ = 414 ps). The extended singlet carbene lifetime of the p-phenyl substituted

phenylcarbene is consistent with the expectation that the zwitterionic, quinoidal singlet

carbene can be stabilized by an electron donating group. The stronger transient

absorption detected for 1BpCH also makes it possible to integrate the carbene signal

intensities (integration of 1PhCH was not performed because the result is imprecise due to

its weak signal intensity).

Figure 3.9. Ultrafast LFP ( λex = 270 nm) of p-biphenyldiazirine in chloroform. (a) The kinetic trace of 1BpCH carbene band monitored at 1570 cm -1 by fitting to an exponential function. (b) The kinetic trace of 1BpCH carbene band obtained by intensity integration.

The integration ( Figure 3.9b) reveals that the concentration of 1BpCH is almost constant from 1 – 20 ps, and then it decays exponentially over 20 ps after the laser pulse.

The lifetime of 1BpCH carbene obtained by fitting the intensity integration is almost the

same with that obtained by fitting the kinetic trace at the peak absorption (1570 cm -1,

Figure 3.9a). However, the near constant concentration before 20 ps reveals that the majority of 1BpCH is formed within 1 ps of the laser pulse. More precise information on

66 the growth time constant of carbene can not be determined because the hot carbene band is so broad that part of the absorption is outside of the spectral window (~ 100 cm -1).

Similar transient absorptions of singlet carbene 1BpCH were also detected in dichloromethane (CH 2Cl 2), cyclohexene (CHE), and methanol-O-d (MeOD), as shown in

Figure 3.10 , Figure 3.11 , and Figure 3.12 , respectively. The decay curves and lifetimes

obtained in these three solvents are shown in Figure 3.13 . The lifetime in CH 2Cl 2 ( τ =

382 ± 32 ps) is shorter than that in CHCl 3 ( τ = 599 ± 37 ps), consistent with the halogen

atom effect reported previously. 45 The significantly shortened lifetimes in CHE ( τ = 38 ±

10 ps) and MeOD ( τ = 19 ± 3 ps), two excellent singlet carbene scavengers, are also consistent with the assignment of the transient IR band to singlet carbenes based on calculations, in further support of the assignment.

67

Figure 3.10. The transient IR spectra of p-biphenylcarbene 1BpCH obtained by ultrafast LFP ( λex = 270 nm) of p-biphenyldiazirine ( p-BpCN 2H) in CH 2Cl 2.

Figure 3.11. The transient IR spectra of p-biphenylcarbene 1BpCH obtained by ultrafast LFP ( λex = 270 nm) of p-biphenyldiazirine ( p-BpCN 2H) in cyclohexene (CHE).

68

Figure 3.12. The transient IR spectra of p-biphenylcarbene 1BpCH obtained by ultrafast LFP ( λex = 270 nm) of p-biphenyldiazirine ( p-BpCN 2H) in methanol-O-d (MeOD).

1 Figure 3.13. The lifetimes of p-biphenylcarbene BpCH obtained by ultrafast LFP ( λex = 270 nm) of p-biphenyldiazirine ( p-BpCN 2H) in (a) CH 2Cl 2, (b) cyclohexene (CHE), and (c) methanol-O-d (MeOD). The decay traces were probed at wavenumbers of maximum absorption and fitted to the exponential function A = A exp(-t/τ) + y0.

69 3.2.5. p-Biphenylmethyldiazirine

p-Biphenylmethyldiazirine ( p-BpCN 2CH 3) in chloroform was studied by ultrafast

LFP ( λex = 270 nm) and the transient spectra of singlet p-biphenylmethylcarbene

1 ( BpCCH 3) so produced are shown in Figure 3.14.

Figure 3.14. Transient spectra generated by ultrafast LFP ( λex = 270 nm) of p- biphenylmethyldiazirine in chloroform. (a) Transient IR spectra showing the growth of 1 BpCCH 3 carbene within time windows 2 – 65 ps. (b) Transient IR spectra showing the 1 decay of BpCCH 3 carbene within time windows 65 – 1763 ps.

1 Again the singlet BpCCH 3 carbene is formed with excess vibrational energy and undergoes VC over 60 ps, then decays with a lifetime of 667 ± 148 ps, monitored at 1585 cm -1 (Figure 3.15 ).

70

Figure 3.15. Ultrafast LFP ( λex = 270 nm) of p-biphenylmethyldiazirine (BpCN 2CH 3) in 1 chloroform. The decay trace of the p-biphenylmethylcarbene BpCCH3 band monitored at 1585 cm -1 was fitted to an exponential function.

1 1 The singlet BpCH and BpCCH 3 carbene lifetimes are both longer than that of

1PhCH (414 ± 74 ps), indicating a para -phenyl kinetic stabilization effect on the singlet

1 1 carbene. As the decay times of relaxed BpCH and BpCCH 3 are similar, we posit that

1,2-H shift does not control the disappearance of these singlet carbenes in chloroform. 46

Calculations ( Figure 3.16 ) predict that the νC=C vibrational mode is one of the most intense bands in all these singlet carbenes. It is also clear that the methyl substituted carbenes are weaker than the parent arylcarbenes; whereas the p-phenyl substitution increased the intensity of the carbene bands.

71

1 1 1 Figure 3.16. Predicted vibrational bands of singlet carbenes PhCH, PhCCH 3, BpCH 1 and BpCCH 3 at the B3LYP/6-31+G(d) level of theory.

3.3. Discussion

Wang et al. previously investigated the photochemistry of p- biphenyldiazomethane and p-biphenyldiazoethane in solution by ultrafast UV –vis

47-49 spectroscopy ( λex = 310 nm). Transient absorption bands with λmax = 360 nm were

1 1 detected and were attributed to singlet carbenes BpCH and BpCCH3. The lifetimes of

1 1 BpCH and BpCCH3 obtained by ultrafast UV-vis spectroscopy in that study are

consistent with the results obtained here with ultrafast IR spectroscopy ( Table 3.5) within

experimental error, even though different techniques, precursors, excitation wavelengths,

solvents, etc were employed. These factors have been known to affect the singlet carbene

lifetimes so obtained. 49 For example, the lifetime of 1BpCH is found to be 599 ± 37 ps in chloroform (CHCl 3) and 382 ± 32 ps in dichloromethane (CH 2Cl 2) with 270 nm excitation by ultrafast IR spectroscopy; previously it was 200 ± 14 ps in acetonitrile

(ACN) with 311 nm excitation, and 220 ± 45 ps (ACN) with 270 nm excitation. Singlet carbene lifetimes are known to be extended by halogenated solvents due to complexation

72 of the empty p orbital of the carbene with non-bonding electron pairs of solvent atoms which account for the data obtained in this study.45 In two reactive solvents, methanol-O-

d (MeOD) and cyclohexene (CHE), the 1BpCH carbene lifetimes are greatly reduced.

Again the lifetimes are consistent with the previous results (cf. Table 3.5). Therefore, the ultrafast IR experiments carried out in this study on the photochemistry of aryldiazirines are consistent with previous ultrafast UV-vis studies. More importantly, vibrational bands are generally narrower than electronic absorptions; and IR bands often can be directly related to specific vibrational modes predicted by calculations, which is not always the case for UV-vis absorptions. For example, singlet phenylcarbene ( 1PhCH) is directly

observed in this study; has not been observed previously by UV-vis spectroscopy because

it is outside of the spectral window (313 nm as predicted by TD-DFT calculations).

73

Carbene Ultrafast UV –Vis Ultrafast IR b c c 200 ± 14 (ACN) , 220 ± 45 (ACN) 599 ± 37 (CHCl 3)

b c 77 ± 5 (CHX) 382 ± 32 (CH 2Cl 2) 1BpCH 27 ± 2 (CHE) b 38 ± 10 (CHE) c 7.9 ± 1.3 (MeOH) b 19 ± 3 (MeOD) c 453 ± 53 (ACN) b, 491 ± 308 (ACN) c 667 ± 148 (CHCl )c 478 ± 59 (ACN) d 3 1 b – BpCCH 3 189 ± 20 (CHX) 170 ± 25 (CHE) b – 10.8 ± 1.1 (MeOH) b – 1 c PhCH – 414 ± 74 (CHCl 3) a All of the experiments were done at room temperature. All ultrafast UV-vis experiments used diazo compounds as the carbene precursors, while the ultrafast IR experiments used diazirine isomers as the carbene precursors. The kinetic traces were probed at the maximum absorptions of each transient bands observed by UV-vis and IR techniques and fitted in exponential functions. b Experiments with 311 nm excitation. c Experiments with 270 nm excitation. d Experiments with 328 nm excitation.

1 1 Table 3.5. The lifetimes (ps) of singlet carbenes BpCH and BpCCH 3 obtained from ultrafast laser flash photolysis with both UV –vis (diazo compound) and IR detection (diazirine). a

3.3. Conclusions

In conclusion, ultrafast IR spectroscopy is a valuable tool for studying the photochemistry of aryldiazirines and allows study of the dynamics of its photoisomerization to diazo compounds and the first direct observation of singlet phenylcarbene. The S 2 state of phenyldiazirine is populated with light of 270 nm

wavelength and decompose directly to both diazo compound and singlet carbene.

3.4. Calculations

DFT and TD-DFT calculations were performed using the Gaussian03 suite of

programs 50 at The Ohio Supercomputer Center. Geometries were optimized at the

B3LYP/6-31+G(d) level of theory with single-point energies obtained at the B3LYP/6-

74 311+G(d,p)//B3LYP/6-31+G(d) level of theory. Vibrational frequency analyses at the

B3LYP/6-31+G(d) level verifed that the stationary points obtained corresponded to energy minima, and were scaled with a factor of 0.9614. The electronic spectra were computed using TD-DFT with Gaussian03 at the B3LYP/6-311+G(d,p) level and using the B3LYP/6-31+G(d) geometry; for these calculations, 10 allowed electronic transitions were calculated.

3.5. Ultrafast Spectroscopy

Ultrafast IR spectroscopic studies were performed using the home-built pump- probe spectrometer described previously.51,52 Samples were prepared in 50 mL of solvent

with unit absorption at the excitation wavelength with 1.0 mm optical length. The entire

set of pump-probe delay positions (cycle) is repeated at least three times, to achieve good

data reproducibility from cycle to cycle. To avoid rotational diffusion effects, the angle

between polarizations of the pump beam and the probe beam was set to the magic angle

(54.7°). Kinetic traces are analyzed by fitting to exponential terms. All experiments were

performed at room temperature.

3.6. Synthesis

The materials, solvents for ultrafast studies were spectrophotometric grade from

Sigma-Aldrich, Inc. and used as received.

Phenyldiazirine 20,40 and 3-phenyl-3-methyl diazirine 4 were prepared according to

the reported procedures. Diazirines should be handled with caution. Explosions were

reported when working with 3-methyldiazirine 53 and 3-n-propyldiazirine 53 as well as when overheating pentamethylenediazirine.53 All operations were carried out in a well-

75 ventilated hood, and caution is taken to avoid exposure to light and heat, and is used immediately after preparation. All of these diazirines were purified by flash column chromatograph on silica gel, with pentane as the eluent. Each fraction was checked in a 1 cm cuvette using a UV-Vis spectrometer and identified by the characteristic diazirine absorption band between 340 – 400 nm. 5,54 Pentane was evaporated using a rotatary

evaporator immersed in a 0 – 4 ºC ice- bath. The colorless liquid obtained contains

about 5 – 20% of pentane as shown by NMR. These phenylalkyl diazirines have a

gasoline-like smell, and are volatile. Further removal of pentane leads to rapid decrease

of yields because diazirines begin to distil. Diazirines are stored at -10 ºC for a few

weeks. Diazirines undergo decomposition on standing and a pink color develops

indicating the isomerization into diazo compounds. NMR spectra were recorded on a

Bruker-400 NMR (FT, 400 MHz, 1H; 100 MHz, 13 C) spectrometer. Chemical shifts ( δ) for 1H, and 13 C are referenced to internal solvent resonances unless stated otherwise.

3.6.1. Phenyldiazirine (PhCN 2H)

A. 2,4,6-triphenyl-1,3,5-triazabicyclo[3.1.0]hexane. Methanol (300 mL) was

cooled to 0 ºC and saturated with for 1 h. The solution was cooled to –10 ºC

with an ice-water bath and saturated with ammonia for 30 min (around 8 – 10 M), and

then cooled to –40 ºC. A solution of t-butyl (17.5 mL) and t-butyl hypochlorite

(17.5 mL, 0.145 mol) was added dropwise via a syringe to the methanolic ammonia

76 solution over a 30 min period. The resulting chloramine solution was warmed to –10 ºC and treated dropwise with 25 mL (0.246 mol) of benzaldehyde. After 1.5 h, the cooling bath was removed and the yellow solution was stored in a freezer overnight. The precipitate that formed was collected and dried under vacuum, yielding 1.56 g of trimer product.

B. Phenyldiazirine .20,40 The trimer (1.6 g, 5 mmol) was suspended in methanol

(10 mL) at 0 ºC and treated with a solution of 0.5 mL of t-butyl hypochlorite in 3 mL of

t-butyl alcohol over a period of 15 min. After stirring at 0 ºC for 1 h, the resulting clear solution was stirred at room temperature for 20 min and then poured into 125 mL of 0.05

M aqueous sodium metabisulfite. The solution was extracted with 10 mL portions of pentane, and the combined extracts were dried, filtered, and evaporated at 0 ºC. The residue was purified by flash chromatography eluting with pentane and monitored by UV

1 spectrometer. A yellow oil of 0.245 g was obtained. λmax = 360 nm (pentane). H-NMR

13 (400 MHz, CDCl 3) δ 7.29–7.35 (m, 3H), 6.89–6.94 (m, 2H), 2.05 (s, 1H). C-NMR (100

MHz, CDCl 3) δ 136.39, 128.31, 128.00, 125.13, 22.36.

O HN NH N N PhCH2NH2 NH2OSO3H Ag2O N NH3

3.6.2. Phenylmethyldiazirine (PhCN 2CH3)

A. N-benzyl-methylphenylketimine . Acetophenone (23.4 mL, 0.2 mol) and

benzylamine (21.8 mL, 0.2 mol) were dissolved in 340 mL of benzene. Then p-

toluenesulfonyl acid monohydrate (1.9 g, 0.01 mol) was added as catalyst. The reaction

77 mixture was refluxed for 24 hr in a Dean–Stark trap. Upon cooling, the precipitates of benzylamine toluenesulfonyl acid salt were removed by filtration. After evaporation of the solvent and drying in vacuum, the product N-benzyl-methylphenylketimine was collected as a slightly yellow oil.

B. Phenylmethyldiazirine .4 N-benzyl-methylphenylketimine (7 g, 33 mmol) in methanol (67 mL) was added dropwise to liquid ammonia (33 mL) and stirred at – 60 ºC for 3 h. A solution of hydroxylamine-O-sulfonic acid (4.67 g, 50 mmol) in methanol (40 mL) was added via an addition funnel and stirred for 2 h. The reaction mixture was allowed to warm up to room temperature and the excess ammonia was allowed to evaporate. The residue was extracted with pentane and concentrated to leave a yellow oil.

This oil was oxidized with freshly prepared (from silver nitrate (8.5 g, 0.05 mol) and sodium hydroxide (2.2 g, 0.053 mol)) in a solution of 1:1 MeOH + H 2O (200 mL) at room temperature for 3 h. The silver salts were filtered out and the product was extracted with pentane, dried over Na 2SO 4, and concentrated to a pale yellow oil.

Chromatography over silica gel with pentane as the eluent afforded 0.45 g of colorless

1 oil. λmax = 370 nm (pentane). H-NMR (400 MHz, CDCl 3) δ 7.36–7.29 (m, 3H), 6.95–

13 6.92 (m, 2H), 1.53 (s, 3H). C-NMR (100 MHz, CDCl 3) δ 139.77, 128.22, 127.37,

125.45, 26.15, 17.64.

3.6.3. p-Bihenyldiazirine ( p-BpCN 2H)

p-Biphenyldiazirine ( p-BpCN 2H) was synthesized following the same procedure

1 as described above for phenyldiazirine. λmax = 375 nm (pentane). H-NMR (400 MHz,

13 CDCl 3) δ 7.30–7.59 (m, 7H), 6.95–7.10 (m, 2H), 2.10 (s, 1H). C-NMR (100 MHz,

CDCl 3) δ 141.01, 140.40, 137.74, 135.41, 128.84, 127.51, 127.05, 125.58, 22.2.

78 3.6.4. p-Bihenylmethyldiazirine ( p-BpCN 2CH 3)

p-Biphenylmethyldiazirine ( p-BpCN 2CH 3) was synthesized following the same

1 procedure as described above for phenylmethyldiazirine. λmax = 375 nm (pentane). H-

13 NMR (400 MHz, CDCl 3) δ 7.30–7.59 (m, 7H), 6.95–7.10 (m, 2H), 1.55 (s, 3H). C-

NMR (100 MHz, CDCl 3) δ 141.01, 140.40, 137.74, 135.41, 128.84, 127.51, 127.05,

125.58, 26.7, 17.91.

3.7. References for Chapter 3

1. Paulsen, S. R. Angew. Chem. 1960 , 72 , 781-782.

2. Schmitz, E.; Ohme, R. Chem. Ber. 1961 , 94 , 2166-2173.

3. Schmitz, E.; Ohme, R. Tetrahedron Lett. 1961 , 612-614.

4. Liu, M. T. H.; Ramakrishnan, K. J. Org. Chem. 1977 , 42 , 3450-3452.

5. Graham, W. H. J. Am. Chem. Soc. 1962 , 84 , 1063-1064.

6. Graham, W. H. J. Org. Chem. 1965 , 30 , 2108.

7. Graham, W. H. J. Am. Chem. Soc. 1966 , 88 , 4677-4681.

8. Graham, W. H. J. Am. Chem. Soc. 1965 , 87 , 4396-4397.

9. Moss, R. A. Acc. Chem. Res. 2006 , 39 , 267-272.

10. Liu, M. T. H. Chemistry of Diazirines ; CRC Press, 1987.

11. Frey, H. M.; Stevens, I. D. R. Proc. Chem. Soc. 1962 , 79-80.

12. Ganzer, G. A.; Sheridan, R. S.; Liu, M. T. H. J. Am. Chem. Soc. 1986 , 108 , 1517- 1520.

13. Turro, N. J.; Butcher, J. A.; Moss, R. A.; Guo, W.; Munjal, R. C.; Fedorynski, M. J. Am. Chem. Soc. 1980 , 102 , 7576-7578.

14. Soundararajan, N.; Platz, M. S.; Jackson, J. E.; Doyle, M. P.; Oon, S. M.; Liu, M. T. H.; Anand, S. M. J. Am. Chem. Soc. 1988 , 110 , 7143-7152.

15. Jackson, J. E.; Soundararajan, N.; Platz, M. S.; Liu, M. T. H. J. Am. Chem. Soc. 1988 , 110 , 5595-5596.

79 16. Amrich, M. J.; Bell, J. A. J. Am. Chem. Soc. 1964 , 86 , 292-293.

17. Rosenberg, M. G.; Kam, S. M.; Brinker, U. H. Tetrahedron Lett. 1996 , 37 , 3235- 3238.

18. Moss, R. A.; Chang, M. J. Tetrahedron Lett. 1981 , 22 , 3749-3752.

19. Bonneau, R.; Liu, M. T. H. J. Am. Chem. Soc. 1996 , 118 , 7229-7230.

20. Smith, R. A. G.; Knowles, J. R. J. Chem. Soc., Perkin Trans. 2 1975 , 686-694.

21. Bayley, H.; Knowles, J. R. Biochemistry 1978 , 17 , 2420-2423.

22. Yamamoto, N.; Bernardi, F.; Bottoni, A.; Olivucci, M.; Robb, M. A.; Wilsey, S. J. Am. Chem. Soc. 1994 , 116 , 2064-2074.

23. Frey, H. M.; Stevens, I. D. R. J. Chem. Soc. 1963 , 3514-3519.

24. Wang, J.; Burdzinski, G.; Kubicki, J.; Platz, M. S.; Moss, R. A.; Fu, X.; Piotrowiak, P.; Myahkostupov, M. J. Am. Chem. Soc. 2006 , 128 , 16446-16447.

25. Zhang, Y.; Kubicki, J.; Wang, J.; Platz, M. S. J. Phys. Chem. A 2008 , accepted for publication .

26. Moss, R. A.; Turro, N. J. In Kinetics and Spectroscopy of Carbenes and Biradicals ; Platz, M. S., Ed.; Plenum: New York, 1990; pp 213f.

27. Schrader, T.; Sieg, A.; Koller, F.; Schreier, W.; An, Q.; Zinth, W.; Gilch, P. Chem. Phys. Lett. 2004 , 392 , 358-364.

28. Cohen, A. D.; Showalter, B. M.; Toscano, J. P. Org. Lett. 2004 , 6, 401-403.

29. Reichardt, C.; Schroeder, J.; Voehringer, P.; Schwarzer, D. Phys. Chem. Chem. Phys. 2008 , 10 , 1662-1668.

30. Kubicki, J.; Zhang, Y.; Wang, J.; Luk, H. L.; Peng, H.-L.; Vyas, S.; Platz, M. S. J. Am. Chem. Soc. 2009 , 131 , 4212-4213.

31. Zhang, Y.; Burdzinski, G.; Kubicki, J.; Platz, M. S. J. Am. Chem. Soc. 2009 , 131 , 9646-9647.

32. Trozzolo, A. M.; Murray, R. W.; Wasserman, E. J. Am. Chem. Soc. 1962 , 84 , 4990-4991.

33. Senthilnathan, V. P.; Platz, M. S. J. Am. Chem. Soc. 1980 , 102 , 7637-7643.

34. Wong, M. W.; Wentrup, C. J. Org. Chem. 1996 , 61 , 7022-7029.

80 35. Cramer, C. J.; Dulles, F. J.; Falvey, D. E. J. Am. Chem. Soc. 1994 , 116 , 9787- 9788.

36. Dorigo, A. E.; Li, Y.; Houk, K. N. J. Am. Chem. Soc. 1989 , 111 , 6942-6948.

37. Roth, H. D. J. Amer. Chem. Soc. 1971 , 93 , 1527-1529.

38. Joines, R. C.; Turner, A. B.; Jones, W. M. J. Am. Chem. Soc 1969 , 91 , 7754-7755.

39. West, P. R.; Chapman, O. L.; LeRoux, J. P. J. Am. Chem. Soc. 1982 , 104 , 1779- 1782.

40. McMahon, R. J.; Abelt, C. J.; Chapman, O. L.; Johnson, J. W.; Kreil, C. L.; LeRoux, J. P.; Mooring, A. M.; West, P. R. J. Am. Chem. Soc. 1987 , 109 , 2456- 2469.

41. McMahon, R. J.; Chapman, O. L. J. Am. Chem. Soc. 1987 , 109 , 683-692.

42. Chapman, O. L.; McMahon, R. J.; West, P. R. J. Am. Chem. Soc. 1984 , 106 , 7973-7974.

43. Warmuth, R.; Marvel, M. A. Chemistry--A European Journal 2001 , 7, 1209-1220.

44. Jackson, J. E.; Soundararajan, N.; White, W.; Liu, M. T. H.; Bonneau, R.; Platz, M. S. J. Am. Chem. Soc. 1989 , 111 , 6874-6875.

45. Wang, J.; Kubicki, J.; Peng, H.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 6604- 6609.

46. Bonneau, R.; Liu, M. T. H.; Kim, K. C.; Goodman, J. L. J. Am. Chem. Soc. 1996 , 118 , 3829-3837.

47. Wang, J.; Burdzinski, G.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2007 , 129 , 2597-2606.

48. Wang, J.; Burdzinski, G.; Gustafson, T. L.; Platz, M. S. J. Org. Chem. 2006 , 71 , 6221-6228.

49. Wang, J., Ph.D. Thesis, the Ohio State University, 2007.

50. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A. J.; T. Vreven; K. N. Kudin; J. C. Burant; J. M. Millam; S. S. Iyengar; J. Tomasi; V. Barone; B. Mennucci; M. Cossi; G. Scalmani; N. Rega, G. A. P., H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J.

81 Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al- Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez, and J. A. Pople. Gaussian 03, Revision E.01 ed.; Gaussian, Inc.: Wallingford CT, 2004.

51. Burdzinski, G.; Hackett, J. C.; Wang, J.; Gustafson, T. L.; Hadad, C. M.; Platz, M. S. J. Am. Chem. Soc. 2006 , 128 , 13402-13411.

52. Wang, J.; Burdzinski, G.; Kubicki, J.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 11195-11209.

53. Schmitz, E.; Ohme, R. Chem. Ber. 1962 , 95 , 795-802.

54. Frey, H. M.; Stevens, I. D. R. J. Chem. Soc. 1963 , 3514-1519.

82 CHAPTER 4

A STYDY OF THE S 1 EXCITED STATE OF PARA -METHOXY 3-PHENYL-3- METHYL DIAZIRINE BY ULTRAFAST TIME RESOLVED UV–VIS AND IR SPECTROSCOPIES AND THEORY

This chapter is reproduced with permission from J. Am. Chem. Soc. 2009 , 131, 13784 . Copyright 2009 American Chemical Society.

4.1. Introduction

Ultrafast time-resolved spectroscopy has been shown to be a powerful tool to

probe the decay dynamics of aryl azide and diazo excited states, and the formation and

decay of nitrene and carbene reactive intermediates, respectively. The singlet excited

states of aryl diazo compounds, 1-5 aryl azides, 6-9 and an excited singlet carbene 10 have recently been detected by ultrafast time-resolved UV–vis and IR spectroscopic studies.

The excited states of aryl diazo compounds typically decay within 300 fs to form singlet arylcarbenes which undergo various intra- and intermolecular reactions. 1,2,10 These

studies also suggest that the excited states of diazo compounds can undergo [1,2]-H shift in concert with nitrogen extrusion, 2 a carbene mimetic reaction that has been termed

“Rearrangement in Excited States” or RIES by Liu et al.11-15 Recent studies of diazo

carbonyl compounds support previous proposals that a portion of the ketene product is

formed from an excited state by an RIES mechanism. 4,16-19 These studies have provided

83 new insights into the very fast processes that proceed on electronically excited state potential energy surfaces.

Calculations have been performed to corroborate these interpretations; however, these studies often focus only on vertical excitation energies with respect to the electronic ground state, without optimization of the structures of the different electronic states being considered. Unfortunately, transient UV–vis absorption experiments do not readily provide direct structural information about excited states comparable to that which can be obtained by ultrafast time-resolved infrared (IR) 20 or Time Resolved Resonance Raman

Spectroscopy. 21 Recent advances in electronic structure calculations (CASSCF, CC2) allow geometry optimizations of excited states and harmonic frequency calculations, analytically or numerically. Therefore, ultrafast time-resolved IR spectroscopy, complemented with modern theoretical methods, appears to be a promising approach for studying excited state dynamics. Qualitative comparisons of calculated singlet electronic excited state vibrational spectra and spectroscopic data have been made previously but are not routine. 22,23 In this study, ultrafast time resolved UV–vis and IR spectroscopies,

combined with modern quantum chemical calculations, were used to study the

electronically excited states of an aryl diazirine.

Diazirines are the cyclic isomers of linear diazo compounds, 24-26 and are often

employed as a preferred source of carbenes due to their greater kinetic stability and ease

of handling. 25-27 Diazirines have also found use as photoaffinity labeling reagents. 28,29

However, the photophysics and photochemistry of diazirines are still active areas of investigation. Recently, we reported studies of arylhalodiazirine photochemistry using ultrafast time-resolved UV–vis spectroscopy (350–360 nm excitation). 30 A transient

84 absorption band was formed within the laser pulse (300 fs). This transient was attributed to a ring-opened zwitterionic species. 30 However, the nature of this zwitterionic species as a discrete reactive intermediate or as a diazirine excited state was not determined.

Scheme 4.1. The proposed zwitterionic species proposed in previous studies.

We recently studied the photochemistry of phenyldiazirine (and the

phenylchlorodiazirine analogue) by ultrafast time-resolved infrared spectroscopy (270 nm

excitation), 31 and found that singlet phenylcarbene and phenyldiazomethane were both

formed in less than 1 ps upon excitation, but the excited states responsible for zwitterion

formation were not characterized.

This area of research was continued, and the first direct observation of the singlet

excited state (S 1) of para -methoxy-3-phenyl-3-methyl diazirine ( p-CH 3OC 6H4CN 2CH 3)

and the formation of the carbene and diazo compound derived from it, using ultrafast

time-resolved IR and UV–vis spectroscopies, was achieved. This diazirine was chosen

for detailed study because of its favorable ground state absorption at 400 nm, which

allows ready promotion to the S 1 state, and, secondly the dipolar nature of S 1 leads this

excited state to a rather long lifetime which facilitates detection.

85 4.2. Ultrafast Spectroscopic Results

4.2.1. Ultrafast UV–vis Spectroscopy with 375 nm Excitation.

Ultrafast laser flash photolysis (LFP) (λex = 375 nm) of p-methoxy-3-phenyl-3- methyl diazirine ( p-CH 3OC 6H4CN 2CH 3) in acetonitrile produces the transient spectra shown in Figure 4.1. A broadly absorbing transient with maximum absorption near 650

nm is formed within the laser pulse (300 fs), and decays with a lifetime of 267 ps ( Figure

4.2, Table 4.1).

Figure 4.1. Transient spectra produced by ultrafast laser flash photolysis ( λex = 375 nm) of p-methoxy-3-phenyl-3-methyl diazirine ( p-CH 3OC 6H4CN 2CH 3) in acetonitrile with a time window of 5–800 ps.

86

Solvents λmax (nm) A τ (ps)

Acetonitrile 650 0.0516 267 ± 19

Chloroform 650 0.0680 271 ± 5

Cyclohexane 620 0.0337 58 ± 1

Methanol 650 0.0370 424 ± 4

Methanol-O-d 650 0.0381 390 ± 10

2,2,2-trifluoroethanol 700 0.0331 846 ± 58 aOnly the long components of the transient decays are tabulated.

Table 4.1. Amplitudes and lifetimes (ps) of transient species produced by ultrafast time- resolved laser flash photolysis (λex = 375 nm) of p-methoxy-3-phenyl-3-methyl diazirine (p-CH 3OC 6H4CN 2CH 3) in selected solvents with UV–vis detection. Lifetimes are obtained by fitting the kinetic traces to an exponential function.a

The lifetime of this transient in chloroform is almost the same ( τ = 271 ps) as in acetonitrile, but is significantly shortened ( τ = 58 ps) in cyclohexane ( Table 4.1). This

observation is consistent with the prior conclusion30 that a very polar intermediate is formed upon photolysis of aryldiazirines. The extended lifetimes of the transient in methanol ( τ = 424 ps) and 2,2,2-trifluoroethanol (TFE) (τ = 846 ps) suggest that the carrier of this transient is not a singlet carbene, because alcohols are well-known singlet carbene quenchers. 5,24

87

Figure 4.2. Ultrafast time-resolved decay curves with UV–vis detection obtained from p- methoxy-3-phenyl-3-methyl diazirine ( p-CH 3OC 6H4CN 2CH 3) with 375 nm excitation. Kinetic traces decay probed at maximum absorbance in (a) cyclohexane, (b) acetonitrile, (c) methanol, (d) 2,2,2-trifluoroethanol, (e) MeOD, and (f) CHCl 3.

A bathochromic shift of the maximum absorption of the transient spectra in polar

and protic solvents ( Figure 4.3) is also observed. A similar observation has been reported

for singlet arylcarbenes and has been attributed to specific solvation of the reactive

intermediate by heteroatoms present in solvent molecules. 3 This result is again consistent with the production of a polar intermediate, a species which is more readily solvated by polar or protic solvents than non-polar solvents. A triplet carbene has biradical rather than

88 zwitterionic character, and a triplet carbene’s lifetime should not be significantly effected much by solvent polarity. Thus, assignment of the transient spectrum to a triplet species seems unwise.

Figure 4.3. Transient UV–vis spectra produced by ultrafast photolysis of p-methoxy-3- phenyl-3-methyl diazirine ( p-CH 3OC 6H4CN 2CH 3). The spectra were generated by ultrafast LFP ( λex = 375 nm) in (a) cyclohexane, (b) methanol, and (c) 2,2,2- trifluoroethanol.

4.2.2. Ultrafast IR Spectroscopy with 400 nm Excitation

Ultrafast laser flash photolysis (LFP) of p-CH 3OC6H4CN 2CH 3 (λex = 400 nm)

produces the transient IR spectra shown in Figure 4.4. A transient vibrational band that absorbs strongly at 1580 cm -1 was observed immediately after the laser pulse (300 fs).

89

Figure 4.4. Transient IR spectra produced by ultrafast photolysis of p-methoxy-3-phenyl- 3-methyl diazirine ( p-CH 3OC 6H4CN 2CH 3) in chloroform. The spectra were generated by ultrafast LFP ( λex = 400 nm) with a time window of 30–1600 ps. The inset shows the kinetics of decay at 1580 cm -1 by fitting to an exponential function.

This transient species decays with a lifetime of 245 ps in chloroform (inset in

Figure 4.4), in good agreement with the decay of the transient absorption ( τ = 271 ps) observed at 650 nm in the same solvent ( Table 4.1). Parent singlet phenylcarbene has

also been detected by ultrafast IR spectroscopy (270 nm excitation) and it has a C=C

vibration at 1580 cm -1 in chloroform. 31 However, the fact that the 1580 cm -1 band decays

with a time constant of 408 ps in methanol-d4 ( Figure 4.5a), much longer than the decay

times recorded in acetonitrile and in chloroform, argues against assigning singlet p-

CH 3OC 6H4CCH 3 carbene as the carrier of this band.

90

Figure 4.5. Ultrafast LFP (λex = 400 nm) of p-methoxy-3-phenyl-3-methyl diazirine ( p- CH 3OC 6H4CN 2CH 3). Kinetic traces probed at maximum absorbance in (a) methanol-d4 (CD 3OD) and (b) cyclohexane fitted to an exponential function.

Furthermore, the decay time constant is consistent with the decay of the transient

species observed by ultrafast time-resolved UV–vis spectroscopy in methanol-O-d

(Figure 4.2e, τ = 390 ps). This convinces us that we have detected the same polar species in both the transient UV –vis and IR experiments. Indeed, the transient 1580 cm -1 IR band decays with a much shorter time constant in cyclohexane (τ = 66 ps, Figure 4.5b),

relative to more polar solvents, which again is in excellent agreement with the decay of

the transient ( τ = 58 ps) observed in this solvent by ultrafast UV–vis spectroscopy

(Figure 4.2a and Table 4.1). This further supports our conclusion that the transient

observed at 640 nm and the IR band at 1580 cm -1 originate from the same intermediate.

Time-dependent density functional theory 32-35 calculations with the B3LYP

functional (TD-B3LYP) and RI-CC2/TZVP calculations predict that the S 1 electronic excited state of this diazirine is initially populated with 350 – 400 nm light ( Table 4.2).

Calculations show that both S2 and S 3 excited states can be reached with 270 nm light.

However, for simplicity, we refer to the state populated with 270 nm light as the S 2 state,

91 because the S 2 and S 3 state are very close in energy, and Kasha’s rule predicts that the S 2 state will be populated rapidly.

TD-B3LYP/6-311+G(d,p)// RI-CC2/ TZVP B3LYP/6-31+G(d) Character Energy/eV Oscillator Character Energy/eV Oscillator State (% contribution) (nm) strength (% contribution) (nm) strength

S1 43 →44 (92 %) 2.93 0.0136 43 →44 (52 %) 3.49 0.0111 (424) 41 →44 (23 %) (355) S2 42 →44 (56 %) 4.43 0.0047 43 →45 (37 %) 4.78 0.0239 41 →44 (36 %) (280) 43 →44 (18 %) (260) S3 41 →44 (41 %) 4.55 0.0124 43 →46 (15 %) 5.56 0.2513 42 →44 (35 %) (272) 43 →45 (28 %) (222) 43 →45 (19 %) 41 →44 (28 %) 43 →44 (19 %) 42 →44 (12 %)

Table 4.2. TD-B3LYP and RI-CC2 vertical excitation energies, oscillator strengths, and the dominant (> 10%) occupied and virtual orbitals contributing to the three lowest energy singlet excitations of p-methoxy-3-phenyl-3-methyl diazirine ( p-

CH 3OC 6H4CN 2CH 3) using the optimized geometry for the S 0 state.

Figure 4.6. Steady state UV–vis absorption spectra of p-methoxy-3-phenyl-3-methyl diazirine ( p-CH 3OC 6H4CN 2CH 3) in pentane.

92 This calculation is in good agreement with the steady state absorption spectra

(Figure 4.6). Therefore, we propose that the polar transient observed by both ultrafast

UV–vis and IR spectroscopies can be assigned to the S 1 excited state of p-

CH 3OC6H4CN 2CH 3. To further support this proposition, optimization of the S 1 excited state of p-CH 3OC6H4CN 2CH 3 at the RI-CC2/TZVP level of theory predicts that the S 1 state has a quinoidal structure in the phenyl group, and the vibrational mode with the strongest intensity is due to aromatic C=C stretches at 1618 cm -1 (Figure 4.7).

Figure 4.7. The optimized geometries of S 0 (left) and S 1 (right) states of p- CH 3OC 6H4CN 2CH 3 at RI-CC2/TZVP level of theory (the bond lengths are in Angstroms).

93

Figure 4.8. Frequencies of the phenyldiazirine electronic excited states (S 1, S 2 and S 3) and ground state S 0 optimized at the RI-CC2/TZVP level of theory.

Computational studies of parent phenyldiazirine also indicate that this aromatic

-1 C=C vibrational band at ~1600 cm can be used as an IR marker for the S 1 state (cf. chapter 1), and interestingly, the S 2 and S 3 excited states and the S 0 ground state are not expected to have a significant band in this region ( Figure 4.8). The reasonable agreement between the experimental data (1580 cm -1) and the predicted value (1618 cm -1) convinces us that it is the S 1 excited state of p-CH 3OC6H4CN 2CH 3 that we observed in this study.

94

Figure 4.9. Dynamics of solvation of the S 1 state of p-methoxy-3-phenyl-3-methyl diazirine ( p-CH 3OC 6H4CN 2CH 3) by ultrafast IR ( λex = 400 nm). (a) cyclohexane; (b) chloroform; (c) methanol-d4.

The complementary experimental and theoretical studies in this work allow us to obtain detailed structural information of the excited states dynamics. For example, the initial S 1 state formed within the laser pulse (300 fs) absorbs at the same frequency at

-1 1572 cm in chloroform, methanol-O-d (CH 3OD) and cyclohexane (Figure 4.9); then

-1 within 20 ps after the laser pulse, the peak undergoes a blue shift of 8 cm in CHCl 3, 12

-1 cm in CD 3OD, while in cyclohexane no shift is observed.

95 The blue-shifting and band narrowing of the newly born transient vibrational absorption is usually observed by ultrafast time resolved IR spectroscopy and assigned to vibrational cooling.36,37 However, in this case, since blue-shifting is not observed in cyclohexane, we instead assign this change to solvation of the S 1 excited state by polar

solvents. This is consistent with the observation of solvation of the polar intermediate

monitored by ultrafast time resolved UV–vis spectroscopy ( Figure 4.3). Thus, the S 1 state of the diazirine is weakly solvated in non-polar solvent (cyclohexane), and more strongly solvated in polar and protic solvents (chloroform, methanol and TFE). This observation is consistent with the polar nature and the quinoidal structure of the S1 excited state. Hence we deduce that there is more double-bond character in the S 1 state in

the more polar solvent. Moreover, from the RI-CC2/TZVP calculations, the S 0 state along

with the vertically excited S 1 and S 2 excited states have dipole moments of 3.51, 6.85, and 3.92 Debyes respectively. The exceedingly large dipole moment of the S 1 state is consistent with a polar intermediate, as expected. The significantly longer lifetime ( Table

4.1) and the large bathochromic shift of the transient visible band ( Figure 4.3) in TFE,

relative to methanol indicates that the S 1 state experiences some specific stabilization by hydrogen bonding with solvent.

96

Figure 4.10. Transient IR spectra produced by ultrafast photolysis of p-methoxy-3- phenyl-3-methyl diazirine in chloroform. The spectra were generated by ultrafast LFP (λex = 400 nm) with a time window of 30–1600 ps. The inset shows the signal growth at 2030 cm -1 by fitted to an exponential function.

This study also allows us to clearly monitor the photoreactions of the S 1 state.

Ultrafast IR ( λex = 400 nm) of p-CH 3OC 6H4CN 2CH 3 in chloroform produces a transient band at 2030 cm -1 ( Figure 4.10 ). The 2030 cm -1 band formed upon excitation is readily

assigned to the isomeric diazo compound. The growth of the diazo compound (274 ± 8

ps, inset in Figure 4.10 ) is nicely correlated with the decay of the S 1 state recorded at

1580 cm -1 ( τ = 245 ± 10 ps, Figure 4.2b). This provides direct evidence that the diazo compound is formed from the S1 state.

4.2.3. Ultrafast IR Spectroscopy with 270 nm Excitation

Previously, we have reported that 270 nm excitation of phenyldiazirine produces phenyldiazomethane and singlet phenylcarbene instantaneously (< 1 ps). 31 Ultrafast time- resolved IR spectroscopy (using 270 nm excitation) of p-CH 3OC6H4CN 2CH 3 diazirine, and the transient IR spectra produced in the 2080 – 1940 cm -1 region are shown in Figure

4.11a.

97

Figure 4.11. Transient IR spectra produced by ultrafast photolysis of p-methoxy-3- phenyl-3-methyl diazirine in chloroform. (a) The spectra were generated by ultrafast LFP (λex = 270 nm) with a time delays of 1 – 200 ps. (b) The growth of the diazo band obtained by fitting the band integration to a biexponential function with the slow component fixed at 240 ps.

The diazo isomer p-CH 3OC6H4CN 2CH 3 is formed with excess vibrational energy

and undergoes vibrational cooling within ~ 50 ps time of the laser pulse. A narrowing and

blue-shift of the diazo band was evident and is attributed to vibrational cooling.31 The

time evolution of the integrated diazo band reveals that it is formed with two time

constants of 4.4 ps and 240 ps with an intensity ratio of 3/1 (Figure 4.11b). The slow

component (240 ps) is in good agreement with the S 1 state lifetime (245 ps) within experimental error and is therefore assigned to the rate of diazo formation from the S 1 state of the diazirine. The fact that 75% of the diazo compound is formed faster than the

S1 state of diazirine decay indicates that there is a second pathway for diazo formation.

Since TD-B3LYP calculations ( Table 4.2) predict that 270 nm excitation can pump p-

CH 3OPhCN 2CH 3 diazirine to the S 2 state, we propose that the fast component of diazo formation originates from the S 2 state, and that the S 2 state decays to the S 1 state as well

as to the diazo compound. We note that the S2 lifetime could be shorter than 4.4 ps, because the diazo band detected before 10 ps is quite broad and a large part of the

98 spectrum falls outside our detection window ( Figure 4.11a). As a result, the band

integration is not very accurate at early delay times (< 10 ps). Therefore, we conclude

that diazo formation is competitive with S 2/S 1internal conversion.

Figure 4.12. Transient IR spectra produced by ultrafast photolysis of p-methoxy-3- phenyl-3-methyl diazirine in chloroform. The spectra were generated by ultrafast LFP (λex = 270 nm) with a time window of (a) 1.5 – 54 ps, and (b) 54 – 3047 ps.

Ultrafast IR spectroscopy with 270 nm excitation of p-CH 3OC6H4CN 2CH 3 in chloroform produces transient absorption in the 1640−1540 cm -1 region, shown in Figure

4.12 . Again, this transient band undergoes vibrational cooling and shifts to 1582 cm -1 within 50 ps of the laser pulse. This band then decays approximately to baseline within 3 ns of the laser pulse.

99

Figure 4.13. Ultrafast IR spectroscopy performed on p-methoxy-3-phenyl-3-methyl diazirine ( p-CH 3OC 6H4CN 2CH 3) in chloroform with 270 nm excitation. (a) Kinetics of carbene decay probed at 1584 cm -1. (b) Kinetics of formation of carbene by fitting the band integrals vs time delay into an exponential function.

The lifetime of this transient, probed at 1584 cm -1, is 1.3 ns (Figure 4.13a). The

1.3 ns lifetime is too long to assign the carrier of the band to the S 1 state, despite the fact

that the S 1 state has been observed in this region ( Figure 4.4) with 400 nm excitation but

with a lifetime of 245 ± 10 ps. Previously, singlet phenylcarbene ( 1PhCH, τ = 414 ps), p-

1 1 biphenylylcarbene ( BpCH, τ = 599 ps), and p-biphenylmethylcarbene ( BpCCH 3, τ =

667 ps) have been observed by ultrafast IR techniques. 31 The carrier of the 1584 cm -1 band is assigned to singlet carbene p-CH 3OC 6H4CCH 3 as the lifetime of this species is within a factor of 3 of other reported singlet carbene lifetimes. Singlet carbene p-

-1 CH 3OC 6H4CCH 3 is predicted to have a strong vibrational mode at 1585 cm ( Table 4.3),

100 in excellent agreement with this assignment. B3LYP calculations predict that the triplet state is the ground state with a small energy gap in the gas phase ( Table 4.4), but the gap is expected to be smaller because this method has been reported to consistently favor the triplet as the ground state. 38 In addition, calculations predict that the singlet is significantly stabilized in polar solvent ( Table 4.3). Therefore intersystem crossing to the triplet state must be slow, or the singlet state may be the ground state in solution, accounting for the extended singlet carbene lifetime.

101

Energy = - 424.07106705 C 0.64639900 -0.95277500 -0.04147500 C -0.74010300 -0.92821900 -0.05245200 C -1.47503500 0.28619300 -0.04155900 C -0.71388000 1.48860800 -0.01481400 C 0.66646100 1.48775300 0.02671900 C 1.35806300 0.26180100 0.00827400 H 1.17244000 -1.90077400 -0.06644100 H -1.27142800 -1.87478300 -0.08602700 H -1.27360000 2.41903300 -0.02207600 H 1.24434000 2.40604600 0.05801800 C -2.90939000 0.44217100 -0.08326900 C -3.73751500 -0.78558900 0.07954200 H -3.27495100 -1.78948700 0.14047400 H -4.31372400 -0.63137600 1.00751300 H -4.50991100 -0.80120300 -0.70325300 C 3.48287800 -0.84128200 0.01471100 H 3.30425700 -1.41886800 -0.90028700 H 4.52408900 -0.51702100 0.04218400 H 3.27204000 -1.46734700 0.89007800 O 2.70739700 0.35297700 0.03322000 Frequency IR intensity Frequency IR intensity Frequency IR intensity (cm -1) (km/mol) (cm -1) (km/mol) (cm -1) (km/mol) 32.7 7.5 824.9 44.9 1416.1 8.1 79.6 1.1 881.1 6.0 1439.8 8.1 102.0 0.9 932.6 0.4 1452.5 5.5 151.6 19.3 966.1 0.6 1464.8 6.7 183.9 0.2 981.5 1.6 1472.5 40.8 227.4 1.7 990.6 21.9 1490.2 15.3 284.6 6.8 1027.4 55.2 1532.2 14.6 307.7 1.4 1050.0 5.9 1584.9 342.4 364.3 0.3 1097.5 17.2 2820.7 99.4 413.7 1.1 1137.8 0.7 2909.1 15.1 489.9 2.7 1157.5 195.0 2923.3 52.5 508.4 6.9 1170.3 14.9 2941.0 11.7 511.4 11.2 1244.4 316.4 2986.3 34.0

Table 4.3. Optimized structure of the singlet state of p-methoxy-phenylmethyl carbene

(p-CH 3OC 6H4CCH 3) at the B3LYP/6-31G(d) level of theory.

102 Solvent ∆H (0 K) ∆G (298 K) gas 4.1 5 cyclohexane 2.8 3.5 chloroform 1.5 2.8 methanol -14.1 -13

Table 4.4. The singlet-triplet energy gap (kcal/mol) of p-methoxy-phenylmethyl carbene (p-CH 3OC 6H4CCH 3) calculated by the PCM model at the B3LYP/6-31G(d) level of theory (positive values indicate a triplet ground state and negative values indicate a singlet ground state).

Finally, and most convincingly, the assignment of the 1584 cm -1 transient (1.3 ns

lifetime) to the singlet carbene is consistent with the fact that methanol, a potent singlet

carbene scavenger, significantly shortens the decay lifetime ( Figure 4.14 ) of the transient

(vide infra ), whereas methanol does not shorten the lifetime of the carrier of the 1580 cm -

1 band, produced by 400 nm excitation assigned to the S 1 state of the diazirine.

Figure 4.14. Ultrafast IR spectroscopy of p-methoxy-3-phenyl-3-methyl diazirine ( p- CH 3OC 6H4CN 2CH 3) in methanol-O-d with 270 nm excitation. A kinetic trace of carbene decay was probed at 1584 cm -1 by fitting into a biexponential function with the slow component fixed at 390 ps.

103 Integration of the carbene band reveals that the singlet carbene has a growth time constant of 3.8 ps ( Figure 4.13b). The carbene growth time constant is very close to that

of the fast component of diazo formation (4.4 ps); therefore, we conclude that the carbene

is formed directly from the S 2 state, even though we acknowledge again that band integration is not very accurate at early delay times. Thus it is likely that the S2 lifetime is

somewhat shorter than 3 ps.

4.3. Discussions

In this study, we report the direct detection of the S 1 state of an aryldiazirine compound p-CH 3OC 6H4CN 2CH 3 by ultrafast time-resolved UV–vis and IR spectroscopies. This assignment is consistent with excited state calculations which predict the aromatic C=C vibrational mode of this S1 state can be used as a diagnostic band. This is the first direct observation of an aryldiazirine S 1 state by UV–vis and IR

spectroscopic methods. The excited states of aryl diazo compounds are reported to

undergo denitrogenation within 300 fs; 1-5 however, the specific excited states involved in

fragmentation were not characterized and it is not clear if the S 1 state or a higher energy

state of the aryl diazo compound was detected. The S 1 states of aryl diazirines have

longer lifetimes (10 − 1000 ps) than the excited states of the isomeric diazo compounds.

The greater kinetic stability of the excited diazirine relative to its diazo counterpart is

reminiscent of the ground state behavior. 39

Some simple dialkyldiazirines are known to exhibit weak fluorescence, 13,40 indicating that the excited states of dialkyldiazirines are intermediates with finite lifetimes in solution, consistent with the finding of this study. The fluorescence lifetime of adamantyldiazirine is on the order of 240 ps at ambient temperature. 40 This

104 fluorescence decay time constant is of the same order as the S 1 state lifetime detected in these ultrafast studies. However, to the best of our knowledge, no fluorescence has been reported for aryldiazirines and we found that the p-CH 3OC6H4CN 2CH 3 compound does

not fluoresce under our experimental conditions to an extent we can detect. The extent to

which arylalkyl diazirine photochemistry resembles that of dialkyldiazirines is unclear.

It is well known that diazirines fragment into carbenes as well as isomerize into

diazo compounds under thermolysis and photolysis conditions. Previously a C=C

stretching band of singlet phenylcarbene at 1580 cm-1 and a diazo stretching band of

phenyl diazomethane at 2030 cm -1 were detected directly by ultrafast time-resolved IR

31 spectroscopic study (λex = 270 nm) of phenyldiazirine. However, the excited state involved was not characterized in that study. In this study, we demonstrate that the decay of the S 1 state of p-CH 3OC6H4CN2CH 3 is correlated with the growth of the diazo compound. Unfortunately a distinct band attributable to the singlet carbene could not be associated with S 1 state decay. Calculations predict that singlet carbene p-

-1 CH 3OC6H4CCH 3 has a C=C vibrational mode at 1585 cm ( Table 4.3). Based on this

result, one could posit that the S 1 state does not form carbene. However, it is likely that the singlet carbene C=C vibrational band is obscured by the S 1 state due to severe

spectral overlap. Indeed, examination of the kinetics at 1580 cm -1 indicates that an offset is present at long delay times in chloroform (Figure 4.4) and cyclohexane ( Figure 4.5b), which is not present in methanol-d4 ( Figure 4.5a), an excellent carbene scavenger.

Therefore we posit that the singlet carbene contributes to the 1580 cm -1 band in chloroform and cyclohexane . In final support of this assignment, we note that the lifetime of 1.2 ns (Figure 4.13a) for singlet carbene p-CH 3OC6H4CCH 3 in chloroform, observed

105 by 270 nm excitation, is consistent with the lifetime of the background absorption observed after the S 1 state, produced by 400 nm excitation, has decayed.

Figure 4.15. Laser flash photolysis ( λex = 355 nm and 308 nm) study of p-methoxy-3- phenyl-3-methyl diazirine ( p-CH 3OC 6H4CN 2CH 3) in chloroform. Peak absorbance of carbene-pyridine ylide at different pyridine concentration.

Another piece of evidence for p-CH 3OC6H4CCH 3 formation from the S 1 state of the diazirine is provided by ns spectroscopy. Solutions of p-CH 3OC6H4CN 2CH 3 containing various concentrations of pyridine were pumped with a 355 nm, 4 ns, laser pulse, and the expected carbene-pyridine ylides were readily detected at 420 nm (Figure

4.15 ). Since TD-B3LYP theory predicts that 355 nm light pumps the p-

CH 3OC6H4CN 2CH 3 diazirine directly to the S 1 state ( Table 4.2), we take this as evidence that some singlet carbene p-CH 3OC 6H4CCH 3 is formed from the S 1 state. However, we acknowledge that the experiments presented in this study can not rule out the possibility that singlet carbene could also be formed via the intermediacy of a diazo compound,

106 which is produced from the initially populated diazirine S 1 state. The detailed mechanism

of carbene formation from the diazirine and diazo compounds warrants further theoretical

which and will be reported later.

Ultrafast IR spectroscopic studies with 270 nm excitation demonstrate that the

diazo compound is formed by two pathways, a fast component (75%) from the S 2 state, and a slow component (25%) from the S 1 state. However, the diazirine S1 state is not

clearly observed by IR spectroscopy with 270 nm excitation. We posit that there is a low

yield of the S1 state and that its vibrational band strength is weak due to the low efficiency of internal conversion and the severe overlap of the two bands (carbene and the

S1 state of diazirine). To test this possibility, methanol-O-d was used to trap the singlet carbene and in this solvent, the transient carbene band at 1584 cm -1 decays bi-

exponentially ( τ1 = 35 ps, τ2 = 390 ps, Figure 4.14 ). The fast component is assigned to carbene decay, consistent with the expectation that methanol solvent rapidly scavenges the carbene, and consistent with the 19 ps lifetime of carbene 1BpCH in the same

31 solvent. The slow component of 390 ps is in excellent agreement with the diazirine S1 state lifetime in CD 3OD ( Figure 4.5a). We propose the following mechanism to explain

the data ( Scheme 4.2).

107

Scheme 4.2. The proposed decay pathways from the excited states of para -methoxy-3- phenyl-3-methyl diazirine.

Ultrafast LFP (375-400 nm) excites the p-CH 3OC6H4CN 2CH 3 diazirine to the S 1 excited state, which isomerizes into the diazo compound in the major decay pathway. A minor pathway is the formation of singlet carbene p-CH 3OC6H4CCH 3. Ultrafast LFP (270 nm) excites to the S 2 electronic state. The direct formation of diazo compound and singlet

carbene is so rapid that it can compete with internal conversion leading to the S 1 state.

4.4. Conclusions

The photophysics and photochemistry of p-CH 3OC6H4CN 2CH3 have been studied using ultrafast time-resolved UV–vis and IR spectroscopies and theoretical methods.

Ultrafast time-resolved laser flash photolysis with 375 nm excitation produced transient absorption bands in the 400 – 700 nm visible region, which are assigned to the S 1 electronic excited state of p-CH 3OC6H4CN 2CH 3 based on modern theoretical studies.

Consistent with this assignment, the strongest vibrational mode of the S 1 excited state of

108 p-CH 3OC 6H4CN 2CH 3, as predicted by calculations, was directly observed in the mid- infrared region. The ultrafast time-resolved spectroscopic experiments also indicate that the S 1 state of p-CH 3OC6H4CN 2CH 3 is a polar intermediate which undergoes solvation

within 20 ps of formation in polar solvents. The ultrafast time-resolved infrared studies

allowed the direct observation of diazo compound and carbene formation from both the

S2 and S 1 states of the diazirines. Carbene formation from the S 1 state was also indirectly

demonstrated by the detection of carbene derived ylides using nanosecond time resolved

laser flash photolysis techniques (355 nm excitation). Ultrafast 270 nm excitation

promotes p-CH 3OC6H4CN 2CH 3 to the S 2 excited state, which decays quickly into the diazo isomer and fragment into a singlet aryl carbene. In addition, about 25% of the isomeric diazo compound is produced from the S 1 state, which, in turn, is produced from

the S 2 state via internal conversion.

4.5. Experimental and Computational Section

4.5.1. Materials

All materials and solvents were purchased from Sigma-Aldrich, Inc. The solvents for ultrafast All materials and solvents were purchased from Sigma-Aldrich, Inc. The solvents for ultrafast studies were spectrophotometric grade obtained from Sigma-

Aldrich, Inc. and used as received.

Synthesis of p-methoxy-3-phenyl-3-methyl diazirine (BpCN 2H). p-Methoxy-3-

phenyl-3-methyl diazirine ( p-CH 3OC6H4CN 2CH 3) was prepared following the method described for 3-phenyl-3-methyl diazirine. 41 Chromatography over silica gel with pentane

109 1 as the eluent afforded yellow solids. λmax = 385 nm (pentane). H-NMR (400MHz,

13 CDCl 3) δ 1.49 (s, 3H), 3.79 (s, 3H), 6.85 (d, 4H, J = 2 Hz). C-NMR (100 MHz, CDCl 3)

δ 159.12, 131.99, 126.66, 113.75, 55.32, 25.89, 17.86.

4.5.2. Ultrafast Spectroscopy

4.5.2.1. Femtosecond Broadband UV–vis Transient Absorption Spectrometer

Transient absorption spectra are recorded with a spectrometer using pump-probe techniques as described elsewhere. 42 The entire set of data are repeated at least three times, to observe data reproducibility from cycle to cycle. Pump pulse energy is about 6

µJ at the sample position. Kinetic traces are analyzed by fitting to a sum of exponential

= − τ + τ terms S(t) ∑ Ai exp( t / i ) C , with independent amplitudes, A i, lifetimes, i, and

offset, C. Convolution with a Gaussian response function is included in the fitting

procedure. The instrument response FWHM is approximately 300 fs. All experiments

were performed at room temperature. Samples were prepared in 50 mL of solvent with

absorption 0.70–0.80 at the excitation wavelength with 1.0 mm optical length. All the

sample solutions were purged with argon prior to the experiments for 5 minutes and

during the experiments.

4.5.2.2. Femtosecond IR Transient Absorption Spectrometer

Ultrafast IR spectroscopic studies were performed using the home-built pump-

probe spectrometer described previously.4,6 Samples were prepared in 50 mL of solvent with unit absorption at the excitation wavelength with 1.0 mm optical length. The entire set of pump-probe delay positions (cycle) is repeated at least three times, to achieve good data reproducibility from cycle to cycle. To avoid rotational diffusion effects, the angle between polarizations of the pump beam and the probe beam was set to the magic angle

110 (54.7°). Kinetic traces are analyzed by fitting to exponential terms. All experiments were performed at room temperature.

4.5.3. Calculations

The geometries of ground state p-methoxy-3-phenyl-3-methyl diazirine ( p-

CH 3OC6H4CN 2CH 3) and singlet carbene p-methoxy phenylmethyl carbene ( p-

CH 3OC6H4CCH 3) were optimized at the B3LYP/6-31G(d) level of theory using hybrid

B3LYP density functional theory 32-35 as implemented in Gaussian 03.43 Vibrational frequency analyses at the same level of theory were utilized to verify that the obtained stationary points corresponded to energy minima. The calculated harmonic vibrational frequencies obtained were corrected with a scaling factor of 0.9614. 44 The electronic

spectra were computed using Time-Dependent Density Functional Theory (TD-DFT) at

the B3LYP/6-311+G(d,p) level using the B3LYP/6-31G(d) geometry; for these

calculations, 10 electronic transitions were calculated. Geometry optimizations and

vertical excitations for p-CH 3OPhCN 2CH 3 were also performed using the second-order coupled cluster method with the resolution-of-the-identity approximation (RI-CC2) method 45-47 as implemented in Turbomole 5.91 .48 The ground state geometry, as optimized at the RI-CC2/TZVP level, was utilized as the starting point in optimization of the singlet excited states. The singlet excited state geometry optimizations were carried out using RI-CC2 methodology with the TZVP basis sets. The stationary points for the singlet excited states were confirmed to be minima by calculating the second derivatives numerically using the NumForce module as implemented in the Turbomole 5.91 suite.

All calculations were performed at The Ohio Supercomputer Center.

111 4.6. References for Chapter 4

1. Wang, J.; Burdzinski, G.; Gustafson, T. L.; Platz, M. S. J. Org. Chem. 2006 , 71 , 6221-6228.

2. Wang, J.; Burdzinski, G.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2007 , 129 , 2597-2606.

3. Wang, J.; Kubicki, J.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 2304-2313.

4. Wang, J.; Burdzinski, G.; Kubicki, J.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 11195-11209.

5. Peon, J.; Polshakov, D.; Kohler, B. J. Am. Chem. Soc. 2002 , 124 , 6428-6438.

6. Burdzinski, G.; Hackett, J. C.; Wang, J.; Gustafson, T. L.; Hadad, C. M.; Platz, M. S. J. Am. Chem. Soc. 2006 , 128 , 13402-13411.

7. Wang, J.; Kubicki, J.; Burdzinski, G.; Hackett, J. C.; Gustafson, T. L.; Hadad, C. M.; Platz, M. S. J. Org. Chem. 2007 , 72 , 7581-7586.

8. Burdzinski, G. T.; Gustafson, T. L.; Hackett, J. C.; Hadad, C. M.; Platz, M. S. J. Am. Chem. Soc. 2005 , 127 , 13764-13765.

9. Kubicki, J.; Zhang, Y.; Wang, J.; Luk, H. L.; Peng, H.-L.; Vyas, S.; Platz, M. S. J. Am. Chem. Soc. 2009 , 131 , 4212-4213.

10. Wang, J.; Kubicki, J.; Hilinski, E. F.; Mecklenburg, S. L.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2007 , 129 , 13683-13690.

11. Jackson, J. E.; Soundararajan, N.; White, W.; Liu, M. T. H.; Bonneau, R.; Platz, M. S. J. Am. Chem. Soc. 1989 , 111 , 6874-6875.

12. White, W. R., III; Platz, M. S. J. Org. Chem. 1992 , 57 , 2841-2846.

13. Modarelli, D. A.; Morgan, S.; Platz, M. S. J. Am. Chem. Soc. 1992 , 114 , 7034- 7041.

14. Huang, H.; Platz, M. S. J. Am. Chem. Soc. 1998 , 120 , 5990-5999.

15. Platz, M. S. Advances in Carbene Chemistry 1998 , 2, 133-174.

16. Burdzinski, G.; Rehault, J.; Wang, J.; Platz, M. S. J. Phys. Chem. A 2008 , 112 , 10108-10112.

17. Burdzinski, G. T.; Wang, J.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 3746-3747.

112 18. Wang, J.; Burdzinski, G.; Kubicki, J.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 5418-5419.

19. Zhang, Y.; Kubicki, J.; Wang, J.; Platz, M. S. J. Phys. Chem. A 2008 , 112 , 11093- 11098.

20. Stoutland, P. O.; Dyer, R. B.; Woodruff, W. H. Science 1992 , 257 , 1913-1917.

21. Garavelli, M.; Negri, F.; Olivucci, M. J. Am. Chem. Soc. 1999 , 121 , 1023-1029.

22. Zgierski, M. Z.; Lim, E. C. J. Chem. Phys. 2005 , 122 , 111103.

23. Wolf, M. M. N.; Schumann, C.; Gross, R.; Domratcheva, T.; Diller, R. J. Phys. Chem. B 2008 , 112 , 13424-13432.

24. Moss, R. A.; Platz, M. S.; Jones, M., Jr. Reactive Intermediate Chemistry ; Wiley: New York, 2004.

25. Liu, M. T. H. Chemistry of Diazirines ; CRC Press, 1987.

26. Moss, R. A. Acc. Chem. Res. 2006 , 39 , 267-272.

27. Jennings, B. M.; Liu, M. T. H. J. Am. Chem. Soc. 1976 , 98 , 6416-6417.

28. Brunner, J.; Senn, H.; Richards, F. M. J. Bio. Chem. 1980 , 255 , 3313-3318.

29. Richard A. G. Smith, J. R. K. J. Am. Chem. Soc. 1973 , 95 , 2.

30. Wang, J.; Burdzinski, G.; Kubicki, J.; Platz, M. S.; Moss, R. A.; Fu, X.; Piotrowiak, P.; Myahkostupov, M. J. Am. Chem. Soc. 2006 , 128 , 16446-16447.

31. Zhang, Y.; Burdzinski, G.; Kubicki, J.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 16134-16135.

32. Becke, A. D. J. Chem. Phys. 1993 , 98 , 5648-5652.

33. Becke, A. D. J. Chem. Phys. 1992 , 96 , 2155-2160.

34. Becke, A. D. J. Chem. Phys. 1992 , 97 , 9173-9177.

35. Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988 , 37 , 785-789.

36. Schrader, T.; Sieg, A.; Koller, F.; Schreier, W.; An, Q.; Zinth, W.; Gilch, P. Chem. Phys. Lett. 2004 , 392 , 358-364.

37. Nibbering, E. T. J.; Elsaesser, T. Appl. Phys. B 2000 , 71 , 439-441.

38. Scott, A. P.; Platz, M. S.; Radom, L. J. Am. Chem. Soc. 2001 , 123 , 6069-6076.

113 39. Liu, M. T. H. Chem. Soc. Rev. 1982 , 11 , 127-140.

40. Buterbaugh, J. S.; Toscano, J. P.; Weaver, W. L.; Gord, J. R.; Hadad, C. M.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 1997 , 119 , 3580-3591.

41. Liu, M. T. H.; Ramakrishnan, K. J. Org. Chem. 1977 , 42 , 3450-3452.

42. Wang, J., Ph.D. Thesis, the Ohio State University, 2007.

43. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A. J.; T. Vreven; K. N. Kudin; J. C. Burant; J. M. Millam; S. S. Iyengar; J. Tomasi; V. Barone; B. Mennucci; M. Cossi; G. Scalmani; N. Rega, G. A. P., H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al- Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez, and J. A. Pople. Gaussian 03, Revision E.01 ed.; Gaussian, Inc.: Wallingford CT, 2004.

44. Scott, A. P.; Radom, L. J. Phys. Chem. 1996 , 100 , 16502-16513.

45. Hattig, C. J. Chem. Phys. 2003 , 118 , 7751-7761.

46. Haettig, C.; Koehn, A. J. Chem. Phys. 2002 , 117 , 6939-6951.

47. Kohn, A.; Hattig, C. J. Chem. Phys. 2003 , 119 , 5021-5036.

48. Ahlrichs, R.; Baer, M.; Haeser, M.; Horn, H.; Koelmel, C. Chem. Phys. Lett. 1989 , 162 , 165-169.

114 CHAPTER 5

ULTRAFAST SPECTROSCOPY OF ARYLCHLORODIAZIRINES: HAMMETT CORRELATIONS OF EXCITED STATE LIFETIMES

This chapter is reproduced with permission from J. Am. Chem. Soc. 2009, 131, 16652. Copyright 2009 American Chemical Society.

5.1. Introduction

Arylhalocarbenes (ArCX), exemplified by PhCF, PhCCl, and PhCBr, have played

an important role in the history of carbene chemistry.1 The aryl group provides a good

UV –vis chromophore; the halogen atom is a π-electron-donor, σ-electron-acceptor substituent, which effectively stabilizes the π-electron-acceptor, σ-electron-donor singlet carbene. 2-4 Therefore, arylhalocarbenes (ArCX) generally have singlet ground states, 5-7 with singlet-triplet energy gaps ( ∆EST ) of ~12 kcal/mol. These carbenes are devoid of

intramolecular rearrangement reactions and intersystem crossing (ISC), which are not a

complicating factor in the studies of alkylcarbenes and arylcarbenes. The ease with which

ArCX species can be generated by photolysis of arylhalodiazirines (1), greatly facilitates

the application of nanosecond laser flash photolysis (ns LFP) experiments since their

widespread introduction almost thirty years ago. 1 Much of our knowledge of singlet

carbenes has been obtained from the intermolecular addition of ArCX to alkenes.

115

Scheme 5.1. Classical photochemical reaction pathways of arylhalodiazirines.

However, the early events in the photophysics and photochemistry of diazirines,

and the mechanism of diazirine excited state bifurcation between carbene and diazo

formation, are still not clearly understood ( Scheme 5.1). Arylhalodiazirines ( 1, X = F, Cl,

Br) have been studied by ultrafast LFP in both acetonitrile (ACN) and cyclohexane

(CHX) and transients have been detected with λmax ≈ 600 – 700 nm and lifetimes of hundreds of picoseconds (ps). The carrier of these transient absorption bands were assigned to the excited states of arylhalodiazirines (1*).8 From the dependence of the

excited state lifetimes on solvent polarity and the electron-donating capacity of X, the

excited states were depicted as diradicals/zwitterions 2.8-11 However, the nature of the excited states (S 1, S 2 or S 3) and their structures were not addressed in detail in previous

studies. 8

Encouraged by our recent progress in studying the excited state (S 1) of an

aryldiazirine with ultrafast UV–vis/IR spectroscopies and ab initio calculations, 12 we

sought to extend our studies by revisiting these unanswered questions regarding the

photochemistry of arylhalodiazirines. Herein, the computational and ultrafast

spectroscopic studies of phenylchlorodiazirine and five other ring-substituted analogues

116

(3) are reported. These studies permit a more precise representation of their excited

states, as well as correlations of the excited state lifetimes with solvent polarity and the

electronic properties of their aryl substituents.

5.2. Ultrafast Spectroscopic Results

Ultrafast LFP ( λex = 375 nm) of phenylchlorodiazirine ( 3a, Y = H) in ACN

affords a broadly absorbing transient with λmax = 625 nm ( Figure 5.1a) within the 300 fs

13 laser pulse. The transient decays biexponentially with τ1 = 2.8 ± 0.3 ps and τ2 = 61.6 ±

7.4 ps ( Figure 5.1b). In keeping with our previous report, 8 this transient is assigned to an excited state of 3a, even though a diradical/zwitterion structure ( 2) has been proposed.

This species is assigned to the S 1 excited state based on calculations (cf. computation section, vide infra ). 14

Figure 5.1. Ultrafast laser flash photolysis ( λex = 375 nm) of phenylchlorodiazirine ( 3a , Y = H) in ACN. (a) Transient spectra of 3a with a time window of 0.3 – 300 ps. (b) Kinetic traces of transient absorptions probed at 640 nm and fitted to exponential functions.

117

We do not assign the 625 nm transient to singlet carbene PhCCl, although this carbene has been reported to have a weak absorption in the ~ 700 nm region, 15,16 and TD

B3LYP/6-31+G(d) calculations predict absorption at 790 nm ( f = 0.0014, Table 5.1); because singlet carbene 1PhCCl (and 1ArCCl in general) decays much more slowly ( τ ≈

1 0.1 – 10 µs) than the transient of Figure 5.1a. The slow component ( τ2 = 61.6 ± 7.4 ps) is assigned to the intrinsic lifetime of the excited state in acetonitrile, in accord with

8 previous report (50 ± 4 ps). We suggest that τ1 represents intramolecular vibrational

relaxation (IVR) of 3* (vide infra ). 17-19

Energy = -729.9408087 Waveleng (nm) f C 1.581537000 1.405375000 0.000051000 790 0.0014 C 0.251479000 1.009135000 0.000016000 318 0.0226 C -0.087328000 -0.368922000 0.000014000 C 0.960740000 -1.322500000 0.000019000 303 0.0002 C 2.295614000 -0.922445000 0.000063000 281 0.3271 C 2.603303000 0.440920000 0.000087000 262 0.0000 H 1.836267000 2.461923000 0.000054000 248 0.0070 H -0.539363000 1.752475000 -0.000009000 H 0.687627000 -2.373679000 -0.000002000 231 0.0054 H 3.090375000 -1.663015000 0.000075000 221 0.0334 C -1.405979000 -0.982957000 -0.000032000 220 0.0000 Cl -2.700793000 0.206442000 -0.000090000 210 0.0410 H 3.642380000 0.761142000 0.000108000

Table 5.1. Singlet phenylchlorocarbene ( 1PhCCl) vertical excitation energies computed at the B3LYP/6-311+G(d,p)//B3LYP/6-31+G(d) level of theory.

118

Five additional arylchlorodiazirines 3, with Y = p-CH 3 ( 3b ), p-CH 3O ( 3c), p-Cl

20 (3d), m-Cl ( 3e), and p-CF 3 ( 3f), were prepared by hypochlorite oxidations of known

arylamidines. 21,22 Ultrafast LFP of these diazirines in ACN all give very similar transients in the 500-700 nm region. These transient spectra and their respective decay curves at

λmax are shown in Figure 5.2 – Figure 5.6. By analogy, these transients are assigned to

the S1 excited state ( 3*) of these diazirines.

Data in ACN show that the lifetimes of the S 1 excited state ( 3* ) are greatly affected by the aryl substituents (Y), in the order τ3c (Y = p-CH 3O, 760.2 ps) > τ3b (Y = p-

CH 3, 150.2 ps) > τ3a (Y = p-H, 61.6 ps) > τ3d (Y = p-Cl, 50.0 ps) > τ3e (Y = m-Cl, 22.1 ps)

> τ3f (Y = p-CF 3, 10.6 ps). It is clear that electron-donating substituents (Y = p-CH 3O, p-

CH 3) increase the S 1 state lifetime and electron-withdrawing substituents (Y = p-Cl, p-

CF 3) shorten the S 1 state lifetime. This trend suggests a positively charged S 1 state ( 3* ).

This result is consistent with previous reports that the lifetimes of 1* (Ar = Ph) increase

with increasing resonance donating ability of X (F > Cl > Br). 8 A noticeable transient

spectral shift ( λmax ) is also observed. Similar to the effect on lifetimes, a bathochromic

shift is observed for the electron-donating substituents, and a hypsochromic shift for the

electron-withdrawing substituents. For example, λmax of the transient spectra for 3c (Y = p-CH 3O) is observed at ~670 nm, and the λmax of 3f (Y = p-CF 3) at ~600 nm, consistent

with charge separation in the S 1 state ( vide infra ).

119

Figure 5.2. Ultrafast laser flash photolysis ( λex = 375 nm) of arylchlorodiazirine ( 3b, Y = p-CH 3) in ACN. (a) Transient spectra of 3b with a time window of 3 – 425 ps. (b) Kinetic traces of transient absorptions were probed at maximum absorptions and fitted to an exponential function.

Figure 5.3. Ultrafast laser flash photolysis ( λex = 375 nm) of arylchlorodiazirine ( 3c , Y = p-OCH 3) in ACN. (a) Transient spectra of 3c with a time window of 27 – 2194 ps. (b) Kinetic traces of transient absorptions were probed at maximum absorptions and fitted to an exponential function.

120

Figure 5.4. Ultrafast laser flash photolysis ( λex = 375 nm) of arylchlorodiazirine ( 3d , Y = p-Cl) in ACN. (a) Transient spectra of 3d with a time window of 1 – 425 ps. (b) Kinetic traces of transient absorptions probed at maximum absorptions and fitted to exponential functions.

Figure 5.5. Ultrafast laser flash photolysis ( λex = 375 nm) of arylchlorodiazirine ( 3e , Y = m-Cl) in ACN. (a) Transient spectra of 3e with a time window of 1 – 98 ps. (b) Kinetic traces of transient absorptions probed at maximum absorptions and fitted to an exponential function.

121

Figure 5.6. Ultrafast laser flash photolysis ( λex = 375 nm) of arylchlorodiazirine ( 3f , Y = p-CF 3) in ACN. (a) Transient spectra of 3f produced with a time window of 1 – 33 ps. (b) Kinetic traces of transient absorptions probed at maximum absorptions and fitted to an exponential function.

Ultrafast LFP study of these six diazirines were also performed in CHCl 3 and

CHX. Similar transients in the visible region are observed, and are attributed to the S 1 state ( 3* ). Decay curves at λmax of all of the transients are shown in Figure 5.7 and

Figure 5.8. In addition to the substituent effects discussed above, the data show that the

S1 lifetimes of 3 also strongly depend on solvent. The S1 state lifetimes increase with

solvent polarity in the order CHX < CHCl 3 < ACN, and λmax of the transients follows the same order. These observations are consistent with a polar S 1 excited state which is

stabilized by an electron-donating para substituent and a polar solvent. These results are explained by computational studies ( vide infra ).

122

Figure 5.7. Ultrafast laser flash photolysis ( λex = 375 nm) of arylchlorodiazirine in chloroform . Kinetic traces of transient absorptions probed at maximum absorption and fitted to exponential functions. (a) 3a , Y = H. (b) 3b , Y = p-CH 3. (c) 3c , Y = p-CH 3O. (d) 3d , Y = p-Cl. (e) 3e, Y = m-Cl. (f) 3f, Y = p-CF 3.

123

Figure 5.8. Ultrafast kinetic traces ( λex = 375 nm) of arylchlorodiazirine in cyclohexane . Kinetic traces of transient absorption probed at maximum absorptions and fitted to exponential functions. (a) 3a , Y = H. (b) 3b , Y = p-CH 3. (c) 3c , Y = p-CH 3O. (d) 3d, Y = p-Cl. (e) 3e , Y = m-Cl. (f) 3f , Y = p-CF 3.

124

5.3. Computational Results

TD-B3LYP/6-311+G(d,p)//B3LYP/6-31+G(d) calculations predict that the S 0 →

S1 vertical transition of 3 (Y = H) is 401 nm (3.09 eV, Table 5.2). This result is

consistent with the broad long wavelength absorption band observed in the steady state

UV–vis spectrum (λmax = 369 nm). Calculations also predict that there is a significant

energy gap (99 nm or 1.01 eV) between the S 1 and S 2 states, and a transition with oscillator strength similar to that of S 0 to S 1. Therefore, we conclude that 375 nm excitation employed in the experiments pumps the molecules to the S 1 state directly, with

little or no intervention of higher excited states (S 2 state and above).

125

Energy = -839.4973716 Waveleng f C -2.227158000 -1.345946000 0.000276000 (nm) C -0.839341000 -1.198057000 0.000111000 401 0.0074 C -0.263231000 0.080793000 0.000080000 C -1.105182000 1.208551000 0.000201000 302 0.0063 C -2.491911000 1.052759000 0.000355000 267 0.0107 C -3.059259000 -0.223679000 0.000392000 240 0.0056 H -2.656434000 -2.344291000 0.000306000 H -0.204606000 -2.076953000 0.000006000 230 0.0000 H -0.684916000 2.208943000 0.000173000 224 0.2137 H -3.127457000 1.934298000 0.000444000 208 0.0002 H -4.139385000 -0.342609000 0.000516000 C 1.210868000 0.287561000 -0.000057000 200 0.1454 N 1.779473000 1.482012000 -0.619396000 196 0.0225 N 1.779653000 1.481616000 0.619914000 195 0.0004 Cl 2.267659000 -1.135099000 -0.000778000

Table 5.2. Chlorophenyldiazirine ( 3, Y = H) vertical excitation energies computed at the B3LYP/6-311+G(d,p)//B3LYP/6-31+G(d) level of theory.

The possible involvement of diazirine excited states in formal carbene processes have intrigued chemists for decade. 5 Numerous product studies and ns LFP spectroscopy studies have shown that at least part of the formally rearranged products of carbenes are produced directly from the diazirine excited states ( Scheme 5.2). 23-27 This carbene mimetic mechanism was named as “rearrangement in the excited state” or RIES by Liu. 28

126

Scheme 5.2. The mechanism of rearrangement in the excited state (RIES) in the photolysis of diazirines.

However, the structure of diazirine excited state involved in RIES is not yet well understood. Calculations have predicted that the UV–vis absorption of parent diazirine

(CH 2N2) and dimethyl diazirine ((CH 3)2CN 2) above 300 nm is due to an n-π* transition, and this n-π* excited state passes through a conical intersection in a nearly barrierless process and leads to a biradical-like structure. 9-11 However, theory has not yet considered

the excited states of asymmetrically substituted diazirines, such as the arylhalodiazirine

(3) of this work. Previously we have studied phenylhalodiazirines by ultrafast LFP

methods and proposed biradical/zwitterionic structures ( 2 in Scheme 5.1) for the

observed transients. 8 However, theoretical studies were not performed. Herein, we have

studied the excited states of phenylchlorodiazirine ( 3, Y = H) in the gas phase at the RI-

CC2/TZVP level of theory. The RI-CC2 computational method has reliably reproduced

other electronic excited state properties.12,29

127

Figure 5.9. S0 (left), S1 (middle), and S 2 (right) of 3* (Y=H) optimized at the RI- CC2/TZVP level of theory; bond lengths are in angstroms.

The ground state of phenylchlorodiazirine ( 3, Y = H) was optimized at the RI-

CC2/TZVP level of theory. Calculations confirmed that the S 1 excited state is an n-π*

9-11 state, consistent with previous results. The first (S 1) and second (S 2) excited states were optimized and confirmed to be minima by computing the analytical second derivatives. Figure 5.9 depicts the ground state (S 0), the first excited state (S 1), and the

second excited state (S 2) geometries of 3. S 1 is not a diradical/zwitterionic species

resembling 2,8 but is instead predicted to be a covalent structure in which the diazirine C-

N bonds have lengthened from 1.468 Å in S 0 to 1.546 – 1.551 Å in S 1. We also observe bond length alternation in S 1: the C2-C3 and C5-C6 bonds shorten, while the other C-C bonds lengthen, giving S 1 somewhat of a quinoidal appearance, consistent with the charge distribution in S 1 ( vide infra ). We propose that the S 1 state of the diazirine can

either directly fragment to produce PhCCl and N 2, or it can relax to the S 0 surface where it can either form the carbene or thermally deactivate.

The Mulliken charge distribution in S 1 of 3 (Y = H) indicates that positive charge

resides on the para carbon (+0.04) and on the diazirine ipso carbon (+0.09), while

128 significant negative charge accumulates on the nitrogen atoms (-0.12, -0.13). Direct interaction of the p-Y substituents of 3 with the positively charged para carbon in S 1 accounts for the observed substituent effects on S 1 lifetimes ( τ2). Similarly, resonance donation from X to the positively charged diazirine carbon of 1* (Ar = Ph) accounts for

8 the previously observed dependence of τ2 on X.

The RI-CC2/TZVP computed dipole moments of the S 0 and S 1 states of 3 (R=H) are 2.28 and 4.78 D, respectively, which explains why the lifetime ( τ2) of S 1 is prolonged

in polar solvents, as observed, where τ2 tracks solvent polarity (CAN > CHCl 3 > CHX).

For any substituent Y, τ2 is longest in ACN and shortest in CHX. The experimental data

agree very well with the polar properties predicted for the S 1 state of diazirine by

calculations.

5.4. Discussion

5.4.1. Doublet Absorption Spectrum

Ultrafast photolysis (λex = 375 nm) of arylchlorodiazirines in solution at room temperature produces transient absorptions with λmax in the visible (500 – 700 nm) region within the time frame of the laser pulses (300 fs). However, we noticed that weak absorptions are also observed in the region for all the arylhalodiazirines (cf.

Figure 5.1 to Figure 5.6), mostly with shoulders at ~ 400 nm. Presumably the λmax occurs outside of the spectral window (400 – 700 nm). It is difficult to obtain accurate spectral and kinetic information of transients in this region, because the intensity is weak, and because the signals overlap with the diazirine ground state absorption (350 – 410 nm, especially for compound 3c (Y = p-CH 3O, cf Figure 5.3). As has been discussed in other

129 chapters, the diazo compound produced via photoisomerization also absorbs here. Both the weak (~400 nm) and the strong (~ 600 nm) absorptions transpire with similar kinetics; and the shapes of transient spectra remain unchanged over their lifetimes. Both bands are shifted to longer wavelengths by electron-donating groups ( Figure 5.1 to

Figure 5.3), and shifted to shorter wavelengths by electron-withdrawing groups ( Figure

5.5 to Figure 5.6). Hence we conclude that both bands are due to the same transient species, namely the S 1 excited state of the diazirine. We note that such two bands spectra

are typical of the S 1 state of other aryldiazirines (e.g. phenyldiazirine, p- biphenyldiazirine), arylalkyldiazirine, 12 and arylhalodiazirines. 8

5.4.2. Biexponential Decay

As shown in Figure 5.1 – Figure 5.8, the kinetic traces probed at λmax of the

transient spectra produced by ultrafast LFP of the arylchlorodiazirines were analyzed by

fitting to a multiexponential function. This method has been extensively used in ultrafast

LFP studies. 30 The transients have either single or biexponential decay. All of the six arylchlorodiazirines ( 3) show biexponential decay in chloroform , while only 3a (Y = H) and 3d (Y = p-Cl) have biexponential decay in acetonitrile and cyclohexane . The fast component ( τ1) has time constants of 0 – 5 ps and account for 20 – 60% of the total spectral intensity. For those transients with only single exponential decay in acetonitrile and cyclohexane , the fast component may exist but are not included by fitting (less than

20% of total intensity). It is noted that this biphasic decay is quite general for the S 1 excited state of the diazirine, and has been reported in aryldiazirines and

8 arylhalodiazirines. The common slow component ( τ2) is assigned to the intrinsic lifetime of S 1. We suggest that the fast component ( τ1) is due to intramolecular vibrational cooling

130

17-19 (IVR). Ultrafast excitation ( λex = 375 nm) of ground state arylhalodiazirine changes its electron density distribution, thus the molecule in the Frank-Condon geometry is not the minimum on the S 1 excited state surface. Therefore we propose that the fast component ( τ1) is associated with structural reorganization from Frank-Condon geometry

to the equilibrium structure. Alternatively, the excitation energy is localized in the

diazirine chromophore upon excitation, and then the excess heat redistributes within the

molecule and produces a thermalized species. This process is known as intramolecular

vibrational cooling (IVR) which generally occurs on the time scale of a few picoseconds

as studied by Rabinovitch and Troe. 17-19

5.4.3. Hammett Correlation

The lifetimes of the S 1 states (slow component, τ2) of p-substituted arylhalodiazirines are tabulated in Table 5.3. Ultrafast LFP studies reveal that the lifetimes ( τ2) of the S 1 states of the arylhalodiazirines ( 3* ) are strongly affected by

substituents (Y), and are consistent with the electron donating ability of substituents. RI-

CC2 calculations predict that the S 1 state has a quinoidal-like structure and positive charges are developed on the diazirine carbon and the para aromatic carbon bearing the

substituent (Y), consistent with ultrafast studies. In order to further understand the

electronic structure of the S 1 state, we studied the Hammett correlations between τ2 and

substituent parameters, in the same manner as ground state phenomena. 1 A Hammett correlation between τ2 and σp is poor, however, we obtain excellent Hammett correlations

+ between τ2 and σp (r = 99.5 – 99.8%) in cyclohexane, acetonitrile, and in chloroform

(Figure 5.10 ).

131

σ + a τ τ τ τ τ Y in 3 p ACN CHCl 3 CHX CH 3OH TFE p-MeO -0.78 760 ± 22 644 ± 78 132 ± 12 95.6 ± 3.8 43.6 ± 3.7

p-CH 3 -0.31 150 ± 8.1 114 ± 8.3 25.8 ± 1.6 136.1 ± 7.9 573.1 ± 93.1 p-H 0.00 61.6 ± 7.4 46.2 ± 5.9 13.3 ± 3.1 59.1 ± 1.8 92.6 ± 8.2 p-Cl 0.11 50.0 ± 5.9 40.6 ± 8.1 8.0 ± 0.8 56.3 ± 3.2 191.3 ± 39.0 m-Cl 0.37 22.1 ± 2.9 16.5 ± 4.0 4.7 ± 0.2 39.3 ± 6.7 53.9 ± 13.0 b p-CF 3 0.53 10.6 ± 0.7 6.9 ± 0.8 2.5 ± 0.2 10.2 ± 0.8 12.8 ± 2.9 a + σp values are from March, J. Advanced Organic Chemistry , Fourth Edn; New York: b + Wiley, 1992, p. 280. σp is taken as identical to σp.

Table 5.3. Lifetimes (ps) of transient absorptions of arylchlorodiazirines obtained by ultrafast laser flash photolysis with 375 nm pulses (300 fs). Only the long components of the transient decays are tabulated.

Furthermore, negative ρ values are obtained, which are consistent with a

positively charged species, stabilized by para substituents. These are among the very few

Hammett correlations of excited state lifetimes. 31-33 Significantly, this study shows that the substituent factors parameterized from ground state structures also apply to the excited state molecules. In Hammett correlations the absolute ρ value is usually a direct indication of the relative amount of charge development in the transition state of a reaction or of an equilibrium. We noticed that the Hammett ρ values vary very little among CHX (-1.27), ACN (-1.37), and CHCl 3 (-1.43), even through the absolute S 1 lifetimes ( τ2) strongly depend on solvent polarity ( Table 5.3). This probably implies that

the main interaction of the S1 state with solvent dipoles occurs at the negative (and sterically unencumbered) diazirine nitrogen atoms, rather than at the para carbon, where the positive charge is rather small and the Y substituent might sterically hinder solvation.

132

+ + Figure 5.10. Hammett correlations for S 1 of 3* : log τ vs σp ; see Table 5.3 for τ and σp . ρ = -1.37 ( r = -0.998) for ACN (black), ρ = -1.43 ( r = -0.995) for CHCl 3 (blue), and ρ = - 1.27 ( r = -0.996) for CHX (red).

5.4.4. Reactions of the S 1 State with Alcohols

Previously we have studied p-methoxy phenylmethyldiazirine ( p-

12 CH 3OC 6H4CN 2CH 3) using ultrafast LFP, and found the that the S 1 state of the diazirine

does not react with methanol or trifluoroethanol (TFE). In addition, hydrogen bonding

interactions between the S 1 state and alcoholic solvents were proposed to explain the

unusually extended S 1 lifetimes in these two solvents. Herein, these arylchlorodiazirines were also studied by ultrafast LFP in methanol and trifluoroethanol (TFE). Similar transients were detected and they are all assigned to the S 1 state. The decay curves were shown in Figure 5.11 and Figure 5.12 and τ2 is tabulated in Table 5.3. Extended

lifetimes were again observed in methanol and TFE for arylhalodiazirines ( 3), except with compound 3c (Y = p-CH 3O), which has significantly shortened lifetimes in these

two solvents ( Table 5.3). In methanol-OD, the lifetime ( τ2 = 131.0 ps, Figure 5.13 ) is

longer than in methanol and there is a kinetic isotope effect (KIE = 1.37). The shorter

133 lifetime in TFE ( τ2 = 43.6 ps) than in methanol ( τ2 = 95.6 ps) indicates that the stronger

acid (TFE) reacts faster with S 1. The KIE and the unexpected low τ2 values indicate that the S 1 excited state of 3c is protonated in MeOH and TFE.

Singlet carbenes are known to react rapidly with alcohols. and significant KIE’s

in these reactions have been reported for singlet p-biphenylyltrifluoromethylcarbene. 34

However, we note that a singlet carbene does not contribute to the transient observed in this ultrafast LFP study. TD-DFT calculations predict that singlet carbenes for the six arylhalodiazirine ( 3) have only one absorption band in the 720 – 860 nm range above 320 nm. Therefore, these results indicate that the transient absorptions observed must be due to the S 1 state and the S 1 state of 3c indeed reacts with alcohols.

134

Figure 5.11. Ultrafast time resolved kinetic traces ( λex = 375 nm) of arylchlorodiazirines in methanol . Kinetic traces of transient absorptions probed at maximum absorptions and fitted to exponential functions. (a) 3a , Y = H. (b) 3b , Y = p-CH 3. (c) 3c , Y = p-CH 3O. (d) 3d , Y = p-Cl. (e) 3e , Y = m-Cl. (f) 3f , Y = p-CF 3.

135

Figure 5.12. Ultrafast time resolved kinetic traces ( λex = 375 nm) of arylchlorodiazirine in trifluoroethanol . Kinetic traces of transient absorptions probed at maximum absorptions and fitted to exponential functions. (a) 3a , Y = H. (b) 3b , Y = p-CH 3. (c) 3c , Y = p-CH 3O. (d) 3d , Y = p-Cl. (e) 3e , Y = m-Cl. (f) 3f , Y = p-CF 3.

136

Figure 5.13. Ultrafast laser flash photolysis ( λex = 375 nm) of arylchlorodiazirine (3c , Y = p-CH 3O) in methanol -OD. Kinetic traces probed at maximum absorption wavelengths fitted to exponential functions.

+ + Figure 5.14. Hammett correlations for S 1 of 3* : log τ vs σp ; see Table 5.3 for τ and σp . ρ = -1.17 ( r = -0.933) for MeOH (green), and ρ = -1.76 ( r = -0.936) for TFE (red).

It is interesting to note that arylchlorodiazirine 3c (Y = p-CH 3O) is the only

diazirine with an S 1 state which reacts with methanol. All the other diazirines behave

normally . We do not find evidence for other diazirines that can react with methanol. The

Hammett plot ( Figure 5.14 ) of the other five diazirines gives a fair linear correlation

137 coefficient in methanol ( r = 93.3%) and TFE ( r = 93.6%). We postulate that the major

interaction of the S 1 state with CHX, ACN, and CHCl 3 involves the negatively charged nitrogen atoms. Hence, this is consistent with the conclusion that the S 1 state of 3c is

protonated by MeOH and TFE.

5.5. Conclusions

In conclusion, experimental and computational studies suggest that 375 nm excitation of arylchlorodiazirines 3 furnishes S 1 excited states with lengthened C-N bonds, positive charge at the para and diazirine carbon atoms, and negative charge at the nitrogen atoms. Such structures rationalize the observed solvent and substituent effects on the excited state lifetimes. Excellent linear relationship in a Hammett plot is obtained

+ between the S 1 state lifetimes and the para substituent parameters σp in cyclohexane, acetonitrile, and chloroform. The chief interaction between solvent and the S 1 state is the negatively charged nitrogen atoms. Hydrogen bond interactions between the S 1 state nitrogen atoms and the alcoholic solvent are proposed to account for the extended S 1 lifetimes in alcohols. However, the S 1 excited state of 3c (Y = p-CH 3O) is protonated in methanol and TFE, and accounts for the unexpected shortened S 1 state lifetimes.

5.6. Experimental Section

5.6.1. Ultrafast Spectroscopy

Ultrafast UV–vis pump-probe absorption measurements were performed using the

home-built spectrometer at the Ohio State University. 13 Solution concentrations were

adjusted to an absorbance of unity in a 1 mm cell. Sample solutions were excited at 375

nm in a stainless steel flow cell equipped with 1 mm thick CaF 2 windows. The entire set 138 of pump-probe delay positions (cycle) was repeated at least three times to observe data reproducibility from cycle to cycle. To avoid rotational diffusion effects, the angle between polarizations of the pump beam and the probe beam was set to the magic angle

(54.7°). Kinetic traces were analyzed by fitting to a sum of exponential terms. All experiments were performed at room temperature.

5.6.2. Calculations

Ground state geometry optimizations were carried out using Becke’s three- parameter hybrid exchange functional with the Lee-Yang-Parr correlation functional

(B3LYP) 35-38 as implemented in Gaussian03 .39 Vibrational frequency analyses were performed to verify that the stationary points obtained corresponded to energy minima.

Frequencies were not scaled. 40 The absolute energies of optimized structures are in

Hartrees. The electronic spectra were computed using Time-Dependent B3LYP (TD

B3LYP) methodology with Gaussian03 at the B3LYP/6-311+G(d,p)//B3LYP/6-31G(d) level. Calculations were also carried out with RI-CC2/TZVP methodology as implemented in Turbomole 5.91 .41-44 All calculations were performed at The Ohio

Supercomputer Center.

5.6.3. Materials

All materials and solvents were purchased from Sigma-Aldrich, Inc. The solvents

for ultrafast studies were spectrophotometric grade from Burdick & Jackson and used as

received. Arylchlorodiazirines are conveniently prepared by Graham’s reaction 20 with known arylamidines. 21,22

139

5.7. References for Chapter 5

1. Moss, R. A.; Turro, N. J. In Kinetics and Spectroscopy of Carbenes and Biradicals ; Platz, M. S., Ed.; Plenum: New York, 1990; pp 213f.

2. Scott, A. P.; Platz, M. S.; Radom, L. J. Am. Chem. Soc. 2001 , 123 , 6069-6076.

3. Irikura, K. K.; Goddard, W. A., III; Beauchamp, J. L. J. Am. Chem. Soc. 1992 , 114 , 48-51.

4. Garcia, V. M.; Castell, O.; Reguero, M.; Caballol, R. Mol. Phys. 1996 , 87 , 1395- 1404.

5. Moss, R. A.; Platz, M. S.; Jones, M., Jr. Reactive Intermediate Chemistry ; Wiley: New York, 2004.

6. Cohen, A. D.; Showalter, B. M.; Toscano, J. P. Org. Lett. 2004 , 6, 401-403.

7. Zhang, Y.; Burdzinski, G.; Kubicki, J.; Platz, M. S. J. Am. Chem. Soc. 2008, 130 , 16134-16135.

8. Wang, J.; Burdzinski, G.; Kubicki, J.; Platz, M. S.; Moss, R. A.; Fu, X.; Piotrowiak, P.; Myahkostupov, M. J. Am. Chem. Soc. 2006 , 128 , 16446-16447.

9. Yamamoto, N.; Bernardi, F.; Bottoni, A.; Olivucci, M.; Robb, M. A.; Wilsey, S. J. Am. Chem. Soc. 1994 , 116 , 2064-2074.

10. Bernardi, F.; Olivucci, M.; Robb, M. A.; Vreven, T.; Soto, J. J. Org. Chem. 2000 , 65 , 7847-7857.

11. Arenas, J. F.; Lopez-Tocon, I.; Otero, J. C.; Soto, J. J. Am. Chem. Soc. 2002 , 124 , 1728-1735.

12. Zhang, Y.; Burdzinski, G.; Kubicki, J.; Vyas, S.; Hadad, C. M.; Sliwa, M.; Poizat, O.; Buntinx, G.; Platz, M. S. J. Am. Chem. Soc. 2009 , 131 , 13784-13790.

13. Burdzinski, G.; Hackett, J. C.; Wang, J.; Gustafson, T. L.; Hadad, C. M.; Platz, M. S. J. Am. Chem. Soc. 2006 , 128 , 13402-13411.

14. Zhang, Y.; Burdzinski, G.; Kubicki, J.; Platz, M. S. Abstracts of Papers, 237th ACS National Meeting, Salt Lake City, UT, United States, March 22-26, 2009 2009 , ORGN-248.

15. Zuev, P. S.; Sheridan, R. S. J. Org. Chem. 1994 , 59 , 2267-2269.

16. Pliego, J. R., Jr.; De Almeida, W. B.; Celebi, S.; Zhu, Z.; Platz, M. S. J. Phys. Chem. A 1999 , 103 , 7481-7486.

140

17. Rabinovitch, B. S.; Rynbrandt, J. D. J. Phys. Chem. 1971 , 75 , 2164-2171.

18. Trenwith, A. B.; Rabinovitch, B. S. J. Chem. Phys. 1986 , 85 , 1694-1695.

19. Schwarzer, D.; Hanisch, C.; Kutne, P.; Troe, J. J. Phys. Chem. A 2002 , 106 , 8019- 8028.

20. Graham, W. H. J. Am. Chem. Soc. 1965 , 87 , 4396-4397.

21. Moss, R. A.; Terpinski, J.; Cox, D. P.; Denny, D. Z.; Krogh-Jespersen, K. J. Am. Chem. Soc. 1985 , 107 , 2743-2748.

22. Kazmirowski, H. G.; Landmann, H.; wALSMANN, P. Pharmazie 1969 , 24 , 378- 380.

23. Jackson, J. E.; Soundararajan, N.; White, W.; Liu, M. T. H.; Bonneau, R.; Platz, M. S. J. Am. Chem. Soc. 1989 , 111 , 6874-6875.

24. White, W. R., III; Platz, M. S. J. Org. Chem. 1992 , 57 , 2841-2846.

25. Modarelli, D. A.; Morgan, S.; Platz, M. S. J. Am. Chem. Soc. 1992 , 114 , 7034- 7041.

26. Huang, H.; Platz, M. S. J. Am. Chem. Soc. 1998 , 120 , 5990-5999.

27. Platz, M. S. Advances in Carbene Chemistry 1998 , 2, 133-174.

28. Bonneau, R.; Liu, M. T. H.; Kim, K. C.; Goodman, J. L. J. Am. Chem. Soc. 1996 , 118 , 3829-3837.

29. Wang, J.; Kubicki, J.; Burdzinski, G.; Hackett, J. C.; Gustafson, T. L.; Hadad, C. M.; Platz, M. S. J. Org. Chem. 2007 , 72 , 7581-7586.

30. Wang, J.; Burdzinski, G.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2007 , 129 , 2597-2606.

31. Harriman, A.; Hosie, R. J. J. Chem. Soc., Faraday Trans. 2 1981 , 77 , 1695-1702.

32. Wagner, P. J.; Truman, R. J.; Scaiano, J. C. J. Am. Chem. Soc. 1985, 107 , 7093- 7097.

33. Sakamoto, M.; Cai, X.; Hara, M.; Fujitsuka, M.; Majima, T. J. Phys. Chem. A 2005 , 109 , 4657-4661.

34. Wang, J.; Kubicki, J.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 2304-2313.

35. Becke, A. D. J. Chem. Phys. 1993 , 98 , 5648-5652.

141

36. Becke, A. D. J. Chem. Phys. 1992 , 96 , 2155-2160.

37. Becke, A. D. J. Chem. Phys. 1992 , 97 , 9173-9177.

38. Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988 , 37 , 785-789.

39. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A. J.; T. Vreven; K. N. Kudin; J. C. Burant; J. M. Millam; S. S. Iyengar; J. Tomasi; V. Barone; B. Mennucci; M. Cossi; G. Scalmani; N. Rega, G. A. P., H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al- Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez, and J. A. Pople. Gaussian 03, Revision E.01 ed.; Gaussian, Inc.: Wallingford CT, 2004.

40. Scott, A. P.; Radom, L. J. Phys. Chem. 1996 , 100 , 16502-16513.

41. Hattig, C. J. Chem. Phys. 2003 , 118 , 7751-7761.

42. Kohn, A.; Hattig, C. J. Chem. Phys. 2003 , 119 , 5021-5036.

43. Haettig, C.; Koehn, A. J. Chem. Phys. 2002 , 117 , 6939-6951.

44. Ahlrichs, R.; Baer, M.; Haeser, M.; Horn, H.; Koelmel, C. Chem. Phys. Lett. 1989 , 162 , 165-169.

142

CHAPTER 6

STUDIES OF THE SUBSTITUTION AND WAVELENGTH EFFECTS ON THE PHOTOCHEMISTRY OF ARYLALKYLDIAZIRINES BY ULTRAFAST TIME- RESOLVED UV–VIS SPECTROSCOPY

6.1. Introduction

Studies in Chapters 2 – 5 indicate that there is a remarkable wavelength effect on the photochemistry of aryldiazirines. The instantaneous formation of phenylcarbene and phenyldiazomethane with 270 nm excitation wavelength light of phenyldiazirine likely proceeds from the initially populated S 2 state, as calculations predict. Direct observation of the S 1 state is achieved with ultrafast laser flash photolysis (LFP, 350 – 400 nm excitations) of p-methoxy phenylmethyldiazirine ( p-CH 3OC 6H4CN 2CH 3). By taking advantage of the relatively long S 1 state lifetime (on the order of a few hundred

picoseconds) of p-CH 3OC 6H4CN 2CH 3, we were able to conclude that a small portion of the S 2 state undergoes internal conversion to the S 1 state when the S 2 state is directly populated. This wavelength-dependent phenomenon is quite rare in organic photochemistry, as most photoreactions occur from the lowest singlet or triplet excited state of the molecule (Kasha’s rule). 1-3

Bogdanova and Popik studied the photochemistry of diazo Meldrum’s acid, 4,5 and reported that product mixtures in the photolysis of diazo Meldrum’s acid depends on the

143 wavelength of irradiation. In this work we report a study of the substitution and wavelength-dependent effects on the photochemistry of arylalkyldiazirines.

6.2. Ultrafast UV–vis Spectroscopies Results and Discussion

6.2.1. Ultrafast LFP (350 nm) of phenyldiazirine

Alkylphenyl diazirines have been utilized as convenient thermal and photochemical precursors of carbenes. 6 We have previously proposed that carbene mimetic 1,2 hydrogen migrations can proceed in diazirine excited states 7 and this process

has been named Rearrangements in the Excited States (RIES). 8-11 In this section, ultrafast

time-resolved UV–vis studies on alkyl substituted aryldiazirines related to this

mechanistic proposal are reported.

Ultrafast LFP ( λex = 350 nm) of phenyldiazirine (PhCN 2H) in acetonitrile generated transient spectra shown in Figure 6.1a, in the spectral window of 400 – 640

nm. A broad transient band centered at 550 nm is observed within the laser pulse (300 fs)

and decays. The transient decay monitored at 550 nm is biexponential with sub

picosecond and multi picosecond components ( τ1 = 0.8 ± 0.2 ps, τ2 = 5.9 ± 1.2 ps, Figure

6.2a). Similar transient absorption was detected in cyclohexane ( Figure 6.1b), yet the transient is weaker and decays faster than in acetonitrile ( Figure 6.2b). The slow component of multi picoseconds is assigned to the intrinsic lifetime of the S 1 state of

phenyldiazirine based on previous studies (cf. Chapter 2 through Chapter 5). Calculations

show that 350 nm light excites the ground state to its lowest excited singlet state. The

short component of the decay of the S1 state is assigned to intramolecular vibrational

relaxation (IVR) of this state. 12,13

144

Figure 6.1. Transient spectra generated by ultrafast LFP (350 nm) of phenyldiazirine in (a) acetonitrile and (b) cyclohexane, at selected time delays.

Figure 6.2. Kinetic traces obtained upon monitoring of the transient spectra generated by ultrafast LFP (350 nm) of phenyldiazirine at 550 nm in (a) acetonitrile and (b) cyclohexane. The kinetic traces were fitted to a bi-exponential function with an instrument response function (IRF) of 300 fs.

145 6.2.2. Ultrafast LFP (350 nm) of 3-methyl and other 3-alkyl-3-phenyldiazirines

The transient absorption of the S 1 state is observed by ultrafast LFP (350 nm) of

3-methyl-3-phenyl diazirine in acetonitrile ( Figure 6.3a). The maximum of transient absorption ( λmax = 580 nm) is red shifted by 30 nm relative to the S 1 state of

phenyldiazirine. Its lifetime ( τ2 = 29.2 ± 5.3 ps) is about 5 times longer than that of the S 1 state of parent phenyldiazirine ( Figure 6.4a) in acetonitrile. In cyclohexane the transient

absorption ( λmax = 525 nm) is shifted to the blue by 55 nm ( Figure 6.3b), and the lifetime

is shortened relative to that in acetonitrile ( Figure 6.4b). These results indicate that

positive charge has developed on the diazirine ipso carbon and is better stabilized, as

expected, by a methyl group as compared to a hydrogen atom.

Figure 6.3. Transient spectra generated by ultrafast LFP (350 nm) of 3-phenyl-3- methyldiazirine in (a) acetonitrile and (b) cyclohexane, at selected time delays.

146

Figure 6.4. Kinetic traces obtained of transient spectra generated by ultrafast LFP (350 nm) of 3-methyl-3-phenyl diazirine at 550 nm in (a) acetonitrile and (b) cyclohexane. The kinetic traces were fitted to a bi-exponential function with an instrument response function (IRF) of 300 fs.

Similar transient absorptions of related S 1 states were also detected by ultrafast

LFP (350 nm) of trideuteromethyl (CD 3, 33% D atom, cf. synthesis section), ethyl, iso- propyl, and tert -butyl substituted arylalkyldiazirines in acetonitrile and cyclohexane, with maximum absorptions ( λmax ) of 550 – 580 nm. The kinetic traces monitored at λmax are shown in Figure 6.5 – Figure 6.8 and the S 1 state lifetimes obtained by fitting to

biexponential functions were tabulated in Table 6.1. Evidently a pronounced solvent effect is observed on the lifetimes of the S 1 state, as the S 1 state lifetime is lengthened in the polar solvent. These trends are also evident upon excitation of parent phenyldiazirine and other alkylphenyl diazirines. The data are consistent with a previous report of Wang et al. with halophenyl diazirines. 14 Following previous studies,14 we conclude that this

state is zwitterionic, although theory predicts that the diazirine ring remains intact in S 1

(cf. Chapter 2).

147

Acetonitrile Cyclohexane

Diazirines τ1 / ps τ2 / ps τ1 / ps τ2 / ps

PhCN 2H 0.8 ± 0.2 5.9 ± 1.2 0.05 ± 0.03 1.2 ± 0.1

PhCN 2-Me 2.1 ± 0.4 29.2 ± 5.3 0.3 ± 0.1 4.8 ± 0.5

PhCN 2-Et 0.9 ± 0.2 20.9 ± 2.1 1.2 ± 0.2 7.8 ± 1.6

PhCN 2-iPr 0.7 ± 0.1 13.0 ± 0.8 1.3 ± 0.2 5.3 ± 1.6

PhCN 2-tBu 3.0 ± 0.4 21.4 ± 2.1 1.6 ± 0.6 14.5 ± 3.1

a PhCN 2-CD 3 1.7 ± 0.3 24.9 ± 4.0 1.0 ± 0.4 7.1 ± 3.3 a The D% purity is 33% as determined from 1H-NMR.

Table 6.1. Lifetimes of various transient absorptions of S 1 states, measured at maximum absorption, produced by ultrafast LFP (350 nm) of alkylphenyldiazirines.

Scheme 6.1. 1,2-Hydrogen shift in the electronically excited state of a diazirine (S 1).

Inspection of Table 6.1 convinces us to attribute the long component of the decay of the S 1 state to a variety of processes (internal conversion to S 0, isomerization to diazo,

carbene formation) and in cyclohexane to 1,2 hydrogen migration to form alkene in

concert with nitrogen extrusion (RIES, Scheme 6.1). The evidence for RIES is clearly

more persuasive in cyclohexane than in the polar solvent. For example, there is a normal

kinetic isotope effect (KIE) on the lifetimes of trideuteromethyl ( Figure 6.8) versus

methylphenyl diazirine in cyclohexane, but no KIE outside of experimental error in

acetonitrile. Furthermore, the lifetime of the tert -butylphenyl diazirine in cyclohexane

148 (Figure 6.7) is quite long relative to that of the other alkylphenyl diazirines in this solvent, but this trend is not observed in acetonitrile. Thus, we conclude that the polarity of the solvent strongly influences the partitioning of the various mechanisms of excited state decay of alkylphenyl diazirines.

Figure 6.5. Kinetic traces obtained of transient spectra generated by ultrafast LFP (350 nm) of 3-ethyl-3-phenyl diazirine at 550 nm in (a) acetonitrile and (b) cyclohexane. The kinetic traces were fitted to a bi-exponential function with an instrument response function (IRF) of 300 fs.

Figure 6.6. Kinetic traces obtained of transient spectra generated by ultrafast LFP (350 nm) of 3-iso -propyl-3-phenyl diazirine (PhCN 2-i-Pr) at 550 nm in (a) acetonitrile and (b) cyclohexane. The kinetic traces were fitted to a bi-exponential function with an instrument response function (IRF) of 300 fs.

149

Figure 6.7. Kinetic traces obtained of transient spectra generated by ultrafast LFP (350 nm) of 3-tert -butyl-3-phenyl diazirine (PhCN 2-t-Bu) at 550 nm in (a) acetonitrile and (b) cyclohexane. The kinetic traces were fitted to a bi-exponential function with an instrument response function (IRF) of 300 fs.

Figure 6.8. Kinetic traces obtained of transient spectra generated by ultrafast LFP (350 nm) of 3-trideuteromethyl-3-phenyl diazirine (PhCN 2CD 3) at 550 nm in (a) acetonitrile and (b) cyclohexane. The kinetic traces were fitted to a bi-exponential function with an instrument response function (IRF) of 300 fs.

6.2.3. Ultrafast LFP of p-biphenyldiazirine with 375, 350 and 275 nm Excitations

Ultrafast photolysis (375 nm) of p-biphenyldiazirine (BpCN 2H) in acetonitrile

produced the transient spectra shown Figure 6.9. Two intense transient absorptions at the

blue edge (410 nm) and the red edge (700 nm) of the detection window are formed at the

earliest delay times that can be monitored after the laser pulse. Within 1 ps of the laser

150 pulse the spectra reshape and narrow due to vibrational cooling and solvation. Finally the two transient absorption bands decay with the same time constant ( τ = 19.2 ps, Figure

6.10a).

Figure 6.9. Transient spectra generated by ultrafast LFP (375 nm) of p-biphenyldiazirine in acetonitrile at selected time delays.

Figure 6.10. Kinetic traces monitored of transient spectra generated by ultrafast LFP (375 nm) of p-biphenyldiazirine at 425 nm in (a) acetonitrile and (b) 2,2,2- trifluoroethanol and fitted to an exponential function.

151 Following the previous discussion of phenyldiazirine, both transients are assigned to the S 1 state of the BpCN 2H diazirine. The solvation process observed is consistent with

the solvent effect on the S 1 lifetime of phenyldiazirine. The characteristic two-band absorption of the S 1 state is reminiscent of that observed for p-methoxy

phenylmethyldiazirine (Chapter 4) 15 and phenylhalodiazirines (Chapter 5). 16 Thus, it

seems reasonable to propose that the S 1 state of phenyldiazirine also has a two-band

absorption pattern, with a band (at < 400 nm) localized outside of the spectral window. In

2,2,2-triluoroethanol, the S 1 lifetime is extended ( τ = 97.0 ps, Figure 6.10b). Similar

results have been observed for studies on p-methoxy phenylmethyldiazirine. 15

Ultrafast photolysis (350 nm) of p-biphenyldiazirine (BpCN 2H) produces similar transient spectra in acetonitrile ( Figure 6.11a), cyclohexane ( Figure 6.11b), and methanol ( Figure 6.11c), and all are assigned to the S 1 state of the diazirine. In

cyclohexane the 400 nm band is also shifted to the blue relative to acetonitrile. The decay

lifetimes of the S 1 state depends on solvent, and it is 13.1 ps in acetonitrile ( Figure

6.12a), 3.0 ps in cyclohexane ( Figure 6.12b), and 17.5 ps in methanol ( Figure 6.12c).

The polar solvent effect is consistent with that of phenyldiazirine.

152

Figure 6.11. Transient spectra generated by ultrafast LFP (350 nm) of p- biphenyldiazirine in (a) acetonitrile (b) cyclohexane and (c) methanol at selected time delays.

153

Figure 6.12. Kinetic traces monitored of transient spectra generated by ultrafast LFP (350 nm) of p-biphenyldiazirine at 400 nm in (a) acetonitrile (b) cyclohexane and (c) methanol at selected time delays and fitted to an exponential function.

Ultrafast LFP (275 nm) of BpCN 2H in acetonitrile produces the transient spectra shown in Figure 6.13 . A transient absorption band centered at 490 nm was formed within the laser pulse (300 fs, Figure 6.13a). Within the laser pulse (300 fs) the 490 nm band rapidly transforms into another band peaking at 430 nm (Figure 6.13b). As the 430 nm

154 band decays, a band centered at 360 nm grows, and an isosbestic point at 375 nm was observed ( Figure 6.13c).

Figure 6.13. Transient spectra generated by ultrafast LFP (275 nm) of p- biphenyldiazirine in acetonitrile at selected time delays.

The 360 nm band is readily assigned to the singlet carbene 1BpCH. Wang et al.

1 detected the transient absorption ( λmax = 360 nm) of singlet carbene BpCH by ultrafast

LFP (310 nm) of p-biphenyldiazomethane in acetonitrile. 17 The carbene lifetime

monitored at 360 nm is 191 ps ( Figure 6.14c), again consistent with previous results ( τ =

155 200 ps). 17 As the singlet carbene decays, the carbene-acetonitrile ylide is observed at 370

nm. The carbene-acetonitrile ylide has a lifetime of 38.6 µs. 17

Figure 6.14. Kinetic traces monitored at (a) 490 nm, (b) 430 nm and (c) 360 nm of transient spectra generated by ultrafast LFP (275 nm) of p-biphenyldiazirine in acetonitrile.

As will be discussed later, we postulate that the carrier of the transient band centered at 430 nm is the S 1 state of the diazirine, based in part on the similarity of the

transient band of the S 1 state at 430 nm produced with light of 375 nm and 350 nm

156 wavelengths. The carrier of the 490 nm band, formed within the laser pulse (300 fs), is the precursor of the S 1 state, and thus is assigned to a higher excited state of diazirine S n

(n > 1). TD-DFT calculations predict that the S 5 state is directly populated with 275 nm

light ( Table 6.2). The lifetime of S n, as monitored at 490 nm, is ~ 290 fs ( Figure 6.14 ).

The S 1 state absorption band ( λmax = 430 nm) decays biexponentially. The short component is ~ 230 fs ( Figure 6.14 ), which is the lifetime of the S n state. The long

component is 4.2 ps, which is assigned to the lifetime of the S 1 state. The lifetimes of the

S1 state depend on the excitation wavelengths: τ = 4.2 ps with λex = 270 nm, τ = 13.1 ps

with λex = 350 nm, and τ = 19.2 ps with λex = 375 nm. We posit that the S 1 state is born

with excess vibrational energy, and its lifetime is shorter. In support of this postulate, we

have shown that phenylcarbene and phenyldiazomethane undergo vibrational cooling

over ~ 100 ps upon ultrafast LFP (270 nm) of phenyldiazirine. 18 However, no vibrational

cooling was observed for the S 1 state and diazo compound with ultrafast LFP (400 nm) of

p-methoxy phenylmethyldiazirine. 15 This is consistent with the current study.

157

Character Energy/eV Oscillator State (% contribution) (nm) strength TD-B3LYP/6-311+G(d,p)//B3LYP/6-31+G(d) a

S1 51  52 (87%) 48  52 (11%) 3.02 (411) 0.0180

S2 48  52 (59%) 49  52 (17%) 51  52 (14%) 4.31 (288) 0.0162

S3 49  52 (69%) 48  52 (18%) 51  53 (13%) 4.35 (285) 0.0442

S4 50  52 (100%) 4.54 (273) 0.0006

S5 51  53 (86%) 49  52 (14%) 4.56 (272) 0.6350 RI-CC2-TZVP b

S1 51  53 (19%) 51  52 (18%) 48  53 (11%) 47  53 (11%) 3.66 (338) 0.0118

S2 49  52 (28%) 49  55 (26%) 51  53 (22%) 4.85 (255) 0.0111

S3 51  54 (37%) 50  52 (24%) 48  54 (13%) 50  56 (10%) 5.02 (247) 0.0001

S4 51  52 (67%) 51  53 (20%) 5.29 (234) 0.8392

S5 51  53 (22%) 47  52 (18%) 48  52 (12%) 51  55 (10%) 5.93 (209) 0.0093 a Using the B3LYP/6-31+G* optimized geometry for the S 0 state. b Using the RI-CC2/TZVP optimized geometry for the S 0 state.

Table 6.2. Vertical excitation energies, oscillator strengths, and the dominant occupied to virtual orbital configurations (>10%) contributing to the three lowest energy singlet excitations of biphenyldiazirine (BpCN 2H) calculated at the TD-B3LYP/6- 311+G(d,p)//B3LYP/6-31+G(d) and RI-CC2/TZVP levels of theory.

158

Figure 6.15. Transient spectra generated by ultrafast LFP (275 nm) of p- biphenyldiazirine in acetonitrile at selected time delays in the spectral window of 400 – 680 nm.

In order to compare the transient spectra of the S 1 state generated by different excitation wavelengths, the transient spectra shown in Figure 6.15 (275 nm) is displayed such that the spectral window is the same as in Figure 6.11 (350 nm). The S 1 state spectra produced by 275 nm excitation are very similar to that produced by 350 nm excitation. The S 1 spectra produced with 275 nm excitation are broader relative to those generated with 350 nm excitation. This is probably due to vibrational cooling because the

S1 state arising from 275 nm excitation has a greater energy content than that produced with 350 nm excitation. In support of this hypothesis, we noted that the diazo band at

-1 2030 cm produced from p-CH 3OC 6H4CN 2CH 3 diazirine with 270 nm wavelength light undergoes vibrational cooling within ~ 60 ps, while the same diazo band shows no VC if p-CH 3OC 6H4CN 2CH 3 diazirine is excited with 400 nm light.

The band intensity of S 1 produced with 275 nm excitation ( Figure 6.15 ) is only

20% – 40% of that produced with 350 nm excitation (Figure 6.11 ), indicating that the

quantum yield of internal conversion from S n  S 1 is ~ 30%. Therefore, ~ 70% of S n

159 undergoes rapid chemical reactions in competition with decay to the S 1 state. In Figure

6.13 , no isosbestic point was observed from the S n to S 1 conversion, also consistent with the proposition that decay pathways other than S 1 state formation proceed from the S n state. Other possible reactions probably include diazo compound formation and, singlet carbene formation. By monitoring the singlet carbene growth at 360 nm, ~ 20 % of the total carbene formation accounts for the estimated portion of singlet carbene produced from the S n state.

Since the S 1 state lifetime is only 4 ps in acetonitrile, the transient absorption observed at 430 nm after 15 ps can not be due to the S 1 state absorption. Therefore, we

posit that the persistent band at 430 nm, which does not decay in 3 ns, is due to a

photoproduct. Ultrafast LFP (270 nm) on the bleached solution of p-biphenyldiazirine

produced a transient band peaking at 430 – 440 nm over 10 – 1000 ps ( Figure 6.16a).

These experiments were performed using same conditions as Figure 6.13 . Transient

absorption with λmax = 435 nm was detected by ultrafast LFP (270 nm) of an authentic sample of p-biphenylcarboxaldehyde ( p-BpCHO) in acetonitrile. Similar transient spectra and kinetics confirm that the 430 – 440 nm band observed at long time delays is largely due to photoproduct p-BpCHO. Even though both p-BpCHO and the S 1 state absorb at

430 nm, the characteristic two-band-absorption of the S 1 state before 15 ps confirms that

the S 1 state is detected, because the p-BpCHO does not absorb at 700 nm ( Figure 6.16 ).

160

Figure 6.16. Transient spectra generated by ultrafast LFP (275 nm) of (a) bleached solution of p-biphenyldiazirine and (b) p-biphenylcarboxaldehyde (p-BpCHO) in acetonitrile at selected time delays.

Ultrafast LFP studies of p-biphenyldiazirine were also performed in methanol

(Figure 6.17 ), cyclohexane ( Figure 6.21 ) and 2,2,2-trifluoroethanol (TFE) with 270 –

280 nm excitation, and similar results were obtained. A transient band assigned to S n

(λmax = 490 nm) is born within the laser pulse (300 fs) and rapidly decays into the S 1 state which has two absorption bands at 420 nm and 690 nm, similar to the S 1 state spectra

obtained with 350 – 400 nm excitation. The S n state has a sub picosecond lifetime and the

S1 lifetime is a few picoseconds. Both the S n and S 1 state fragment to form the singlet

carbene with λmax = 360 nm. The singlet carbene lifetime is greatly reduced in methanol

(τ = 10.5 ps, Figure 6.17 ), and in cyclohexane ( τ = 38 ps, Figure 6.21 ), relative to acetonitrile ( τ = 191 ps), consistent with previous studies by Wang et al .17

161

Figure 6.17. Kinetic traces monitored at (a) 490 nm, (b) 420 nm and (c) 360 nm of the transient spectra generated by ultrafast LFP (270 nm) of p-biphenyldiazirine in methanol.

162

Figure 6.18. Kinetic traces monitored at (a) 490 nm, (b) 420 nm and (c) 360 nm of transient spectra generated by ultrafast LFP (270 nm) of p-biphenyldiazirine in cyclohexane.

6.2.4. Ultrafast LFP of p-biphenylmethyldiazirine with 375, 350, 310, and 270 nm

Excitations.

Ultrafast LFP (375 nm) of BpCN 2CH 3 in acetonitrile generated the transient spectra shown in Figure 6.19 . Transient absorption bands peaking at 420 nm and 690 nm

are produced within the laser pulse (300 fs), then reshape and narrow within 1 ps of the

163 laser pulse, possibly due to vibrational cooling and solvation. The same time constants, within experimental error, are obtained for the transient decay monitored at 410 nm and

690 nm, indicating that both transient bands are associated with a common species, the S 1 state of the diazirine based on previous studies. The S 1 state lifetime is 28.3 ps in acetonitrile ( Figure 6.20a), and 35.5 ps in methanol ( Figure 6.20b).

Figure 6.19. Transient spectra generated by ultrafast LFP (375 nm) of p- biphenylmethyldiazirine in acetonitrile at selected time delays.

164

Figure 6.20. Kinetic traces monitored at 410 nm of transient spectra generated by ultrafast LFP (375 nm) of p-biphenylmethyldiazirine in (a) acetonitrile and (b) methanol.

Transient spectra produced upon ultrafast LFP (350 nm) of BpCN 2CH 3 in

acetonitrile are shown Figure 6.21 . In addition to the transient absorptions which are

attributed to the S 1 state (the characteristic two-band absorption pattern at 410 nm and >

650 nm produced by 375 nm excitation), another transient band peaking at 490 nm is also formed within the laser pulse (300 fs). As the 490 nm band rapidly decays over 1 ps, the

S1 state band grows and undergoes spectral reshaping due to solvation. The band centered at 490 nm is assigned to the S n excited state which forms the S 1 state via internal

conversion. TD-DFT calculations predict that 350 nm light pumps both the S 1 and S 2 states of BpCN 2CH 3 ( Table 6.3). Therefore, we postulate that both the S 1 and S 2 states are populated with 350 nm excitation and the S 2 state is the carrier of the 490 nm band.

The lifetime of the S 2 state is 260 fs as probed at 490 nm ( Figure 6.22a). The S 1 state lifetime is 25.2 ps as monitored at 400 nm (long component, Figure 6.22b); the short

component ( τ1 = 0.6 ps) is attributed to solvation.

165

Figure 6.21. Transient spectra generated by ultrafast LFP (350 nm) of p- biphenylmethyldiazirine in acetonitrile at selected time delays.

166

Figure 6.22. Kinetic traces monitored at (a) 490 nm and (b) 400 nm of transient spectra generated by ultrafast LFP (350 nm) of p-biphenylmethyldiazirine in acetonitrile.

167

Energy = -650.3184119 Wavelength f C 0.001894000 1.199820000 -0.295676000 (nm) C 1.393408000 1.220895000 -0.300019000 411 0.023 C 2.135836000 0.064080000 -0.014475000 C 1.427078000 -1.114664000 0.276536000 290 0.042 C 0.036417000 -1.130403000 0.280752000 283 0.057 C -0.710169000 0.024790000 -0.005263000 272 0.583 H -0.541479000 2.106075000 -0.547760000 H 1.902155000 2.149346000 -0.538732000 269 0.001 H 1.964257000 -2.027464000 0.511669000 260 0.004 H -0.480473000 -2.051076000 0.536138000 254 0.002 C 3.621697000 0.086065000 -0.020012000 N 4.304267000 -0.909156000 0.854494000 231 0.009 N 4.303917000 -1.199670000 -0.340785000 230 0.023 C -2.194186000 0.004125000 -0.001247000 229 0.010 C -2.931096000 1.099749000 0.480300000 C -2.902351000 -1.111932000 -0.479018000 224 0.003 C -4.324920000 1.080651000 0.483427000 221 0.069 H -2.405320000 1.962377000 0.880246000 214 0.045 C -4.296161000 -1.131994000 -0.475108000 H -2.354535000 -1.959643000 -0.881162000 213 0.022 C -5.014180000 -0.035571000 0.005901000 210 0.007 H -4.873358000 1.936374000 0.868519000 209 0.021 H -4.822182000 -2.003002000 -0.857148000 H -6.100675000 -0.050917000 0.008749000 205 0.025 C 4.377376000 1.354215000 -0.328366000 205 0.006 H 4.140546000 1.723999000 -1.333220000 203 0.014 H 4.136002000 2.145709000 0.391207000 199 0.004 H 5.453913000 1.171058000 -0.280870000

Table 6.3. Calculations of p-biphenylmethyldiazirine (BpCN 2CH 3) at the TD-B3LYP/6- 311+G(d,p)//B3LYP/6-31+G(d) level of theory.

Ultrafast LFP (275 nm) of BpCN 2CH 3 in acetonitrile generated the transient spectra shown in Figure 6.23 . A transient absorption with λmax = 490 nm is generated within the laser pulse (300 fs) and is assigned once again to the S 2 excited state. Then, the

S2 state rapidly transforms into the S 1 state with absorptions of λmax = 410 nm over 1 ps.

168 The S 1 state decays over 50 ps. The transient absorptions observed over 50 – 3000 ps at

420 nm and 520 nm are due to photoproducts ( vide infra ).

The lifetime of the S 2 state monitored at 490 nm is 336 fs ( Figure 6.24a). The S 1 state monitored at 425 nm has a short lifetime component of 2.3 ps and a long component

1 of 25.3 ps ( Figure 6.24b). p-Biphenylmethylcarbene ( BpCCH 3) produced from

photolysis of p-biphenyldiazoethane has been reported to have a lifetime of 453 ps with absorption at 360 nm. 19 However, in this study the decay time constant obtained at 360

1 nm is only 68.7 ps. The discrepancy in the BpCCH 3 lifetimes is caused by the strong

fluorescence of photoproduct 4-vinylbiphenyl (VB) at 360 nm. The 420 nm band is due

to re-excitation of photoproduct (BpCOCH 3) and the 515 nm band is due to re-excitation of 4-vinylbiphenyl (VB). This was determined by ultrafast LFP of a photo bleached solution of BpCN 2CH 3 diazirine in acetonitrile (Figure 6.25a). There are two transient

absorption bands at 430 nm and 520 nm. The carrier of the absorption band at 430 nm is

confirmed to be 4-acetylbiphenyl (BpCOCH 3, Figure 6.25b), and that of the 520 nm band was reported to be characteristic of a substituted biphenyl (Bp) by Wang et al .19

169

Figure 6.23. Transient spectra generated by ultrafast LFP (275 nm) of p- biphenylmethyldiazirine in acetonitrile at selected time delays.

170

Figure 6.24. Kinetic traces monitored at (a) 490 nm, (b) 415 nm and (c) 360 nm of transient spectra generated by ultrafast LFP (275 nm) of p-biphenylmethyldiazirine in acetonitrile.

171

Figure 6.25. Transient spectra generated by ultrafast LFP (275 nm) of (a) a bleached solution of p-biphenylmethyldiazirine and (b) BpCOCH 3 in acetonitrile at selected time delays.

172

Figure 6.26. Kinetic traces monitored at (a) 495 nm, (b) 413 nm and (c) 360 nm of transient spectra generated by ultrafast LFP (275 nm) of p-biphenylmethyldiazirine in cyclohexane.

Excited states of biphenyldiazomethane (BDM*) and biphenyldiazoethane

(BDE*) have been previously observed by Wang et al. with reported transient absorption

17,19 with λmax = 490 nm. Surprisingly, the transient spectra of the S 2 excited states of both p-biphenyldiazirine and p-biphenylmethyldiazirine were also detected with λmax = 490 nm and the spectra are very similar to the excited states of the diazo compounds. This

173 leads to an interesting question: do the isomeric diazirine and diazo compounds share the same excited states? However, it is noted that the S 2 state of BpCN 2H and BpCN 2CH 3 diazirines decay into the S 1 states of the respective diazirines, whereas the Sn excited states of the BpCN 2H and BpCN 2CH 3 diazo compounds fragments to form their

respective singlet carbenes. The different reaction pathways argue against a common

excited state for both diazo/diazirine isomers, unless very different regions of a common

excited state surface are populated. One possible explanation is that both of the excited

states of the diazirines and diazo compounds excited with 270 – 310 nm light are ππ *

states, largely localized on the biphenyl group, and thus they share similar electronic

structure and spectra. Indeed, TD-DFT and RICC2 calculations support the proposition

that excited states populated by 270 – 310 nm light resides on the biphenyl group ( Figure

6.27 ).

It is also possible that the excited states of the biphenyldiazirine and its diazo

isomer are very similar but not common to both species. In support of this point of view,

RI-CC2/TZVP calculations on phenyldiazirine and phenyldiazomethane (cf. chapter 2)

predict that the S 3 state of the diazirine is a ring-open structure, and the dinitrogen moiety

is bent out-of-plane in the S3 state of isomeric phenyldiazomethane, resembling the S 3 state of the diazirine. The additional phenyl group in diazirines BpCN 2H and BpCN 2CH 3 lowers the energy of these excited states to the extent that they now can be pumped by

270 nm excitation, unlike phenyldiazirine. Calculations indeed predict that 270 nm can excite up to the S4 state for BpCN 2H and BpCN 2CH 3 diazirines.

174

S1 state S2 state S3 state (isocontour value ±0.005) (isocontour value ±0.001) (isocontour value ±0.003)

Figure 6.27. Excited state difference density plots for the S 1, S 2 and S 3 states relative to the ground state electron density for p-biphenyldiazirine, as calculated at the RI- CC2/TZVP level of theory. A red surface surrounds areas where electron density is depleted after vertical excitation from S 0; a green surface surrounds areas where electron density is accumulated in the excited state.

6.3. Conclusions

In this work, phenylalkyldiazirines were studied by ultrafast time resolved UV– vis spectroscopy and the S 1 excited state lifetimes of diazirines produced with 350 nm

light were found to dependent on substitution of the alkyl groups, as well as the solvent

polarity. The evidence for 1,2-H migration from the S 1 excited states was more compelling in the nonpolar solvent cyclohexane than in acetonitrile. Wavelength- dependent photochemistry of p-biphenyldiazirine and p-biphenylmethyldiazirine was studied with 270 – 375 nm excitation. The S 1 excited state was observed by monitoring its transient absorptions at 410 nm and 690 nm, produced with 375 nm excitation. When higher excitation energies (270 nm) were used, the S 2 excited state was observed by monitoring its 490 nm absorption band. The S 2 excited state of BpCN 2CH 3 diazirine

decays into the S 1 excited state over 300 fs, and into singlet carbene at 360 nm. These

remarkable wavelength-dependent results do not conform to the Kasha’s rule , and are

consistent with the large energy separation predicted between the S 1 and S 2 excited states of diazirines (cf. Chapter 2).

175

6.4. Ultrafast Spectroscopy

Ultrafast IR spectroscopic studies were performed using the home-built pump- probe spectrometer described previously.20,21 Samples were prepared in 50 mL of solvent

with unit absorption at the excitation wavelength with 1.0 mm optical length. The entire

set of pump-probe delay positions (cycle) is repeated at least three times, to achieve good

data reproducibility from cycle to cycle. To avoid rotational diffusion effects, the angle

between polarization of the pump beam and the probe beam was set to the magic angle

(54.7°). Kinetic traces are analyzed by fitting to exponential terms. All experiments were

performed at room temperature.

6.5. Synthesis

All materials and solvents were purchased from Sigma-Aldrich, Inc. The solvents

for ultrafast studies were spectrophotometric grade from Sigma-Aldrich, Inc. and used as

received.

Phenyldiazirine 22,23 and 3-phenyl-3-methyl diazirine 24 were prepared according to

literature procedures and have been described in chapter 3. 3-Phenyl-3-ethyl diazirine, 3-

phenyl-3-iso -propyl diazirine, and 3-phenyl-3-t-butyl diazirine were prepared by the

same method used to prepare 3-phenyl-3-methyl diazirine. 24 Diazirines should be handled with caution. Explosions were reported when working with 3-methyldiazirine 25 and 3-n-propyldiazirine 25 as well as when overheating pentamethylenediazirine.25

176 3-Phenyl-3-ethyl diazirine. λmax /nm: 375 (pentane). δH ( 400 MHz, CDCl 3): 0.88 (t, 3H,

J = 7.6 Hz), 2.07 (q, 2H, J = 7.6 Hz), 6.98 (m, 2H), 7.28-7.39 (m, 3H). δC (100 MHz,

CDCl 3): 139.06, 128.25, 127.32, 125.58, 29.97, 23.26, 8.53.

3-Phenyl-3-isopropyl diazirine. λmax /nm: 375 (pentane). δH (250 MHz, CDCl 3): 0.70 (d,

6H, J = 7.0 Hz), 2.79 (m, 1H), 6.83-6.92 (m, 2H), 7.11-7.25 (m, 3H).

3-Phenyl-3-tert -butyl diazirine. λmax /nm: 375 (pentane). δH (400MHz, CDCl 3): 0.94 (s,

9H), 7.09-7.34 (m, 5H). δC (100 MHz, CDCl 3): 136.69, 128.91, 128.71, 128.10, 35.72,

33.61, 27.74.

3-Phenyl-3-trideuteromethyl diazirine. λmax /nm: 375 (pentane). δH (400MHz, CDCl 3):

1.56 (m, 2H, 33% D atom), 6.82 (d, 2H, J = 7.2 Hz), 7.32-7.37 (m, 3H).

6.6. References for Chapter 6

1. Turro, N. J. Modern Molecular Photochemistry ; University Press: Menlo Park, CA, 1978.

2. Turro, N. J.; Ramamurthy, V.; Cherry, W.; Farneth, W. Chem. Rev. 1978 , 78 , 125-145.

3. Kasha, M. Discuss. Faraday Soc. 1950 , No. 9 , 14-19.

4. Bogdanova, A.; Popik, V. V. J. Am. Chem. Soc. 2003 , 125 , 1456-1457.

5. Bogdanova, A.; Popik, V. V. J. Am. Chem. Soc. 2003 , 125 , 14153-14162.

6. Liu, M. T. H. Chemistry of Diazirines ; CRC Press, 1987.

7. Nigam, M.; Platz, M. S.; Showalter, B. M.; Toscano, J. P.; Johnson, R.; Abbot, S. C.; Kirchhoff, M. M. J. Am. Chem. Soc. 1998 , 120 , 8055-8059.

8. Bonneau, R.; Liu, M. T. H.; Kim, K. C.; Goodman, J. L. J. Am. Chem. Soc. 1996 , 118 , 3829-3837.

9. Jackson, J. E.; Soundararajan, N.; White, W.; Liu, M. T. H.; Bonneau, R.; Platz, M. S. J. Am. Chem. Soc. 1989 , 111 , 6874-6875.

10. White, W. R., III; Platz, M. S. J. Org. Chem. 1992 , 57 , 2841-2846.

177 11. Modarelli, D. A.; Morgan, S.; Platz, M. S. J. Am. Chem. Soc. 1992 , 114 , 7034- 7041.

12. Rabinovitch, B. S.; Rynbrandt, J. D. J. Phys. Chem. 1971 , 75 , 2164-2171.

13. Charvat, A.; Abmann, J.; Abel, B.; Schwarzer, D.; Henning, K.; Luther, K.; Troe, J. Phys. Chem. Chem. Phys. 2001 , 3, 2230-2240.

14. Wang, J.; Burdzinski, G.; Kubicki, J.; Platz, M. S.; Moss, R. A.; Fu, X.; Piotrowiak, P.; Myahkostupov, M. J. Am. Chem. Soc. 2006 , 128 , 16446-16447.

15. Zhang, Y.; Burdzinski, G.; Kubicki, J.; Vyas, S.; Hadad, C. M.; Sliwa, M.; Poizat, O.; Buntinx, G.; Platz, M. S. J. Am. Chem. Soc. 2009 , 131 , 13784-13790.

16. Zhang, Y.; Wang, L.; Moss, R. A.; Platz, M. S. J. Am. Chem. Soc. 2009 , 131 , 16652-16653.

17. Wang, J.; Burdzinski, G.; Gustafson, T. L.; Platz, M. S. J. Org. Chem. 2006 , 71 , 6221-6228.

18. Kubicki, J.; Zhang, Y.; Wang, J.; Luk, H. L.; Peng, H.-L.; Vyas, S.; Platz, M. S. J. Am. Chem. Soc. 2009 , 131 , 4212-4213.

19. Wang, J.; Burdzinski, G.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2007 , 129 , 2597-2606.

20. Burdzinski, G.; Hackett, J. C.; Wang, J.; Gustafson, T. L.; Hadad, C. M.; Platz, M. S. J. Am. Chem. Soc. 2006 , 128 , 13402-13411.

21. Wang, J.; Burdzinski, G.; Kubicki, J.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 11195-11209.

22. McMahon, R. J.; Abelt, C. J.; Chapman, O. L.; Johnson, J. W.; Kreil, C. L.; LeRoux, J. P.; Mooring, A. M.; West, P. R. J. Am. Chem. Soc. 1987 , 109 , 2456- 2469.

23. Smith, R. A. G.; Knowles, J. R. J. Chem. Soc., Perkin Trans. 2 1975 , 686-694.

24. Liu, M. T. H.; Ramakrishnan, K. J. Org. Chem. 1977 , 42 , 3450-3452.

25. Schmitz, E.; Ohme, R. Chem. Ber. 1962 , 95 , 795-802.

178 CHAPTER 7

ULTRAFAST TIME-RESOLVED UV–VISIBLE AND INFRARED SPECTROSCOPIC STUDIES ON STYRYLCARBOMETHOXY CARBENE

This chapter is reproduced with permission from J. Am. Chem. Soc. 2009 , 131, 13602. Copyright 2009 American Chemical Society.

7.1. Introduction

Singlet carbenes are reactive intermediates that undergo unique inter- and

intramolecular reactions. 1,2 The cyclopropanation reaction of a singlet carbene and an olefin is only one example and has been frequently studied both experimentally and theoretically. Historically, the stereospecific carbene-alkene addition mechanism has been one of the most important criteria in determining the spin multiplicity of a carbene.3

Thus a vinylcarbene which contains a carbene center in conjugation with a C=C double bond will undergo a rapid cyclization reaction ( Scheme 7.1). Nevertheless, vinylcarbenes

are of theoretical significance in carbene chemistry because the conjugation with a C=C

double bond results in a delocalized carbene. Considerable attention has been paid to

vinylcarbenes since the synthesis of from vinylcarbenes and

vinylcarbenoids was discovered by Closs and coworkers in early 1960s. 4 However, few direct characterizations of vinylcarbenes were available at the start of this work.

179 Arylcarbenes can be regarded as a special class of vinylcarbenes where the C=C double bond resides within aromatic ring. Extensive experimental studies have been reported with singlet arylcarbenes, 5,6 whose intermolecular reactions are often faster than intersystem crossing to their triplet ground states in solution. 1 Ring-expansion of phenylcarbene has been reported in gas phase, 7 matrixes 8-11 and in the inner phase of a

hemicarcerand. 12 The vinylcarbene-cyclopropene rearrangement has been proposed to

account for the phenylcarbene-cycloheptatetraene rearrangement. 13 However, this

reaction is not known in solution (cf Chapter 2).

Scheme 7.1. The intramolecular reactions of vinylmethylenes.

Vinylcarbenes are generally more reactive and thus more difficult to detect than arylcarbenes. Parent vinylmethylene was detected in its triplet ground state in rigid matrixes by electron spin resonance (ESR) spectroscopy, 14 and found to exist in two

geometric isomers, cis and trans ( Scheme 7.1). 14 Such cis /trans geometric isomerism has

also been reported for triplet naphthylcarbene 15 and vinyl radical. 16 The more reactive singlet vinylmethylenes, have been invoked as the key intermediates leading to cyclopropenes via ring closure. 4,17-20 However, parent singlet vinylmethylene has eluded

direct detection to date. Singlet vinylmethylene has been predicted to be 12 – 13 kcal/mol

180 higher than the ground-state triplet in previous MCSCF, CIS, and DFT calculations, and has a highly delocalized allylic diradical structure. 21-25

Sheridan and Zuev 26 spectroscopically characterized several singlet ground-state vinylchlorocarbenes produced upon photolysis of vinylchlorodiazirines isolated in cryogenic matrixes. Vinylchlorocarbenes were found to be stable for several days at 9 –

25 K and can be photochemically converted to cyclopropenes. 26 Subsequent studies by the Moss and Sheridan groups of vinylchlorocarbenes involved nanosecond (ns) time- resolved laser flash photolysis (LFP) techniques in solution. 27 They found that the cyclopropene product is formed cleanly upon photolysis of vinylchlorodiazirines, and that vinylchlorocarbenes can be intercepted with alkenes or pyridine. The lifetimes of vinylchlorocarbenes were deduced to be 20 – 40 ns in pentane. 27 These spectroscopic

investigations of singlet vinylchlorocarbenes are greatly facilitated by halogen

substitution, not only because the intersystem crossing from the initially formed singlet

ground state to triplet is now energetically unfavorable, but also because the cyclization is

retarded by stabilizing of the singlet state. Comparable studies on a singlet vinylcarbene

with a low-lying triplet state are still lacking, and remain a challenging application of

ultrafast time-resolved techniques.

Vinyldiazomethane has been utilized as the photolytic precursor of vinylcarbenes

in previous cryogenic studies. 4,14 Decomposition of a vinyldiazo compound should, in principle, produce a singlet vinylcarbene. However, it was known that 3-diazoalkenes undergoes spontaneous 1,5-cyclization to produce 5H-pyrazoles at room temperature,28 complicating laser flash photolysis (LFP) studies in solution. Vinyldiazoacetates

(Scheme 7.2) are stable at room temperature and have been used to produce

181 vinylcarbenoids in asymmetric cyclopropanation 29,30 . Methyl styryldiazoacetate 31

(PhCH=CHCN 2CO 2CH 3, 1, Scheme 7.2) was chosen for this inaugural ultrafast (fs/ps) time-resolved study, not only because the phenyl group provides a good UV–vis chromophore for absorbing light, but also because the ester moiety provides a good IR probe. The anticipated styryl carbomethoxy carbene PhCH=CHCCO 2CH 3 ( 2) is predicted

by DFT calculations to have triplet ground state ( vide infra ). 32

Scheme 7.2. Proposed photochemical reactions of PhCH=CHCN 2CO 2CH 3 upon photolysis.

7.2. Ultrafast Time-resolved Spectroscopic Studies

7.2.1. Ultrafast Time-resolved UV–vis Study with 1 in Acetonitrile

Ultrafast laser flash photolysis (LFP, λex = 310 nm) with 1 in acetonitrile (ACN)

produced the transient spectra shown in Figure 7.1. A broad transient in the spectral window (340 – 600 nm) is formed within the laser pulse (300 fs). As it decays, a new transient is observed with λmax = 385 nm. The former band is attributed to an excited state

33 of the diazo precursor ( 1* ) based on previous studies. The latter transient ( λmax = 385 nm) is attributed to singlet vinylcarbene ( 12). TD-B3LYP calculations predict that 12 absorbs at 377 nm ( f = 0.1220, Figure 7.2), in excellent agreement with the experimental results.

182 As the singlet carbene ( 12) decays, two weak bands at 350 nm and 440 nm are observable over 3 ns ( Figure 7.1). These two weak bands are attributed to triplet

vinylcarbene 32. This assignment is again consistent with TD B3LYP calculations

(Figure 7.3) which predict triplet carbene 32 absorptions at 342 nm ( f = 0.5761), 392 nm

(f = 0.0028), and 457 nm ( f = 0.0001), in good agreement with the experimental data.

Figure 7.1. Transient UV–vis spectra generated by ultrafast time-resolved LFP ( λex = 310 nm) of PhCH=CHCN 2CO 2CH 3 in acetonitrile within time windows (a) 0.5 – 1.0 ps and (b) 2 – 1300 ps.

183 TD B3LYP spectrum 0.5 310.222 0.45 0.4 0.35 0.3 0.25 0.2 377.441 0.15

Oscillator Strength (f) 0.1 0.05 0 300 320 340 360 380 400 420 440 460 480 500 520 540 560 580 600 Wavelength / nm

1 1 Figure 7.2. Simulated absorption spectrum of singlet carbene PhCH=CHCCO 2CH 3 ( 2) in the gas phase. The vertical transition energies (nm) and oscillator strength ( f) are shown as lines from the TD-B3LYP/6-311+G(d,p)//B3LYP/6-31G(d) calculatins. The overall absorption bands are obtained by Lorentzian broadening.

TD B3LYP spectrum 0.7 341.784 0.6

0.5

0.4

0.3 318.855 0.2

Oscillator Strength (f) 301.478 0.1 391.504 456.671 0 300 320 340 360 380 400 420 440 460 480 500 520 540 560 580 600 Wavelength / nm

3 3 Figure 7.3. Simulated absorption spectrum of triplet carbene PhCH=CHCCO 2CH 3 ( 2) in the gas phase. The vertical transition energies (nm) and oscillator strength ( f) are shown as lines from the TD-B3LYP/6-311+G(d,p)//B3LYP/6-31G(d) calculations. The overall absorption bands are obtained by Lorentzian broadening.

1 The absorption band of 2 monitored at 385 nm ( λmax ) decays exponentially with a lifetime of 25.7 ± 1.6 ps (Figure 7.4a). This is a particularly short singlet carbene lifetime 34,35 in ACN solvent and we will attribute this to rapid intramolecular cyclization

to form cyclopropene ( 3, Scheme 7.2) on the basis of time-resolved IR spectral data ( vide

infra ). The decay of singlet carbene ( 12) is not accompanied by the growth of transient

184 UV–vis absorption that is attributable to cyclopropene 3. This is not surprising because

this molecule is predicted to absorb at 327 nm ( f = 0.0251), which is outside of our spectral window ( Figure 7.5). Triplet carbene ( 32) is formed from the singlet state ( 12)

via intersystem crossing. However, the kinetics of the growth of triplet carbene can not be

accurately determined because of severe spectral overlap with the more intense singlet

carbene band.

185

Figure 7.4. The kinetic traces monitored at 390 nm obtained by ultrafast time-resolved LFP ( λex = 310 nm) of PhCH=CHCN 2CO 2CH 3 ( 1) in (a) acetonitrile, (b) cyclohexane, and (c) chloroform. The kinetic traces were fitted to exponential functions.

186 TD B3LYP spectrum

236.414 0.2

0.15 217.919

0.1

0.05

Oscillator Strength (f) 327.22 267.404281.567 0 200 250 300 350 400 450 500 550 600 Wavelength / nm

Figure 7.5. Simulated absorption spectrum of cyclopropene (3) in the gas phase. The vertical transition energies (nm) and oscillator strength (f) are shown as lines from the TD-B3LYP/6-311+G(d,p)//B3LYP/6-31G(d) calculations. The overall absorption bands are obtained by Lorentzian broadening.

Figure 7.6. Transient UV–vis spectra generated by ultrafast time-resolved LFP ( λex = 310 nm) of PhCH=CHCN 2CO 2CH 3 in chloroform within time windows (a) 0.3 – 4.8 ps and (b) 5 – 600 ps.

187 Similar transient absorptions of singlet carbene 12 are also observed in chloroform

(Figure 7.6) and cyclohexane. In these two solvents the singlet carbene decays bi-

exponentially ( Figure 7.4). We postulate that the fast component (< 1 ps) is due to

intramolecular vibrational relaxation (IVR). 36 The slow component is the relaxed singlet

1 1 carbene 2 lifetime. The lifetime of 2 in chloroform ( τ2 = 33.3 ± 2.5 ps) is similar to that

obtained in acetonitrile. However, in cyclohexane, the lifetime is only 6.5 ± 1.1 ps.

Sharply reduced singlet carbene lifetimes in a nonpolar solvent relative to acetonitrile,

have been reported with singlet arylcarbene, Wolff rearrangement reactions, thus this

pattern has precedant. 37

The weak and broad absorption in the 500 – 600 nm range over long time scales

(400 ps to 3 ns in Figure 7.1b and Figure 7.6b) is attributed to the re-excitation of photoproduct. This proposition is verified through a bleach experiment in which a fresh solution of PhCHCHCN 2CO 2CH 3 (1) in acetonitrile (A = 1.0 at 310 nm in a 1 mm cuvette) was irradiated with 308 nm laser pulses for 10 minutes until the diazo precursor

(1) has completely disappeared (absorbance due to 1 at 310 nm reduced to < 0.1); then ultrafast time-resolved LFP experiments were performed with this bleached solution and the transient spectrum so produced is shown in Figure 7.7. The data demonstrate that the broad transient absorption band observed ~ 500 nm originates from photoproduct produced from primary photolysis.

188

Figure 7.7. Transient spectra produced by LFP ( λex = 310 nm) of bleached solution of PhCHCHCN 2CO 2CH 3 ( 1) in acetonitrile.

7.2.2. Ultrafast Time-resolved UV–vis Study in Methanol

Similar transient spectra were obtained by ultrafast LFP (λex = 310 nm) of (1) in

methanol ( Figure 7.8) and methanol-OD ( Figure 7.9). In methanol, the excited state of the diazo precursor ( 1* ) is observed in the visible region and it decays within the instrument response time (300 fs) to form singlet carbene 12 with an absorption maximum at 385 nm. However, the transient absorption of singlet carbene ( 12) undergoes reshaping within 3 ps after the laser pulse, shown as the slight growth and redshift of the maximum absorption peak. This is due to vibrational cooling (VC), consistent with prior ultrafast UV-vis studies of carbene esters. 38,39 The relaxed singlet carbene ( 12) decays with a lifetime of 38.4 ± 2.4 ps in methanol (Figure 7.10a). The triplet carbene ( 32) is

again observed at 340 nm and 440 nm. The triplet carbene yield in methanol is lower than

that in acetonitrile and chloroform, consistent with expectations that methanol intercepts

singlet carbene in competition with intersystem crossing to the triplet carbene. It is noted

that VC is slower in methanol-OD (~ 20 ps, Figure 7.9) relative to methanol. The

189 observation that VC is faster in alcoholic solvents than in non-protic solvents, is consistent with past reports. 40

Figure 7.8. Transient UV–vis spectra generated by ultrafast time-resolved LFP ( λex = 310 nm) of PhCHCHCN 2CO 2CH 3 in methanol in time windows (a) 1.3 – 3.2 ps and (b) 9 – 118 ps.

190

Figure 7.9. Transient UV–vis spectra generated by ultrafast time-resolved LFP ( λex = 310 nm) of PhCHCHCN 2CO 2CH 3 in methanol-OD in time windows (a) 2 – 13 ps and (b) 22 – 353 ps.

Figure 7.10. The kinetic traces monitored at 390 nm obtained by ultrafast time-resolved LFP of PhCHCHCN 2CO 2CH 3 (1) with 310 nm light in (a) methanol and (b) methanol- OD. The kinetic traces were fitted to a single exponential function.

191 In methanol-OD, the lifetime of the singlet carbene is 91.1 ± 4.4 ps ( Figure

7.10b). The kinetic isotope effect (KIE = 2.37) demonstrates that the carrier of transient absorption reacts with the OH bond of the solvent, exactly what one would expect for a singlet carbene, supporting the assignment of the transient spectra ( λmax = 390 nm) to singlet carbene ( 12). The KIE effect is also consistent with previous proposition by

Schmitz 41 that a cation is formed by singlet carbene abstraction of a proton from the

solvent and the stable products from the delocalized cations were formed ( Scheme 7.3).

TD-B3LYP calculations predict that the cation absorbs at 347 nm ( Figure 7.11 ).

However, no time-resolved spectroscopic evidence for a significant yield of a cation was

obtained in our study, presumably because of its low yield.

41 Scheme 7.3. Photochemical reactions of PhCH=CHCN 2CO 2CH 3 proposed by Schmitz.

192 TD B3LYP spectrum

1 347.359

211.969 Oscillator Strength (f) 225.286 263.85 390.916 456.114 0 200 250 300 350 400 450 500 550 600 Wavelength / nm

Figure 7.11. Simulated absorption spectrum of cation in the gas phase formed by protonation of the singlet carbene. The vertical transition energies (nm) and oscillator strength ( f) are shown as lines from the TD-B3LYP/6-311+G(d,p)//B3LYP/6-31G(d) calculations. The overall absorption bands are obtained by Lorentzian broadening.

The observation that the relaxed singlet carbene (12) decays at almost the same rate in methanol as in acetonitrile runs counter to most studies of singlet carbenes. 34,35,42

We speculate that the singlet carbene lifetime in acetonitrile is unusually short because of cyclopropene (3) formation, rather than control by intersystem crossing and/or intermolecular reaction, as is usually the case for the more thoroughly studied aryl carbenes.33,38,42 The data indicate that the rate of cyclization in methanol is slower than in acetonitrile, probably because of more intimate solvation of the carbene (cf. Section 7.3. computational section). 43 The shortened singlet carbene lifetime ( τ = 6.5 ± 1.1 ps) in the non-coordinating solvent cyclohexane is consistent with this interpretation.

7.2.3. Ultrafast Time-resolved IR Studies in Chloroform

Ultrafast time-resolved photolysis ( λex = 270 nm) of PhCH=CHCN 2CO 2CH 3 (1)

in chloroform produces transient IR spectra (Figure 7.12 ) in the diazo stretching region.

The negative band observed at 2083 cm -1 is due to the bleaching of the precursor diazo

band (N=N). A broad positive band is observed on the lower wavenumber side within 1

193 ps of the laser pulse. Then this band undergoes blue-shifting and narrowing within 40 ps; at the same time the bleaching band at 2083 cm -1 recovers its partial intensity (31%). The

behavior is typical for transient species born with excess vibrational energy, and

undergoes vibrational cooling (VC) over tens of picoseconds.44 The spectral evolution

ceases 40 ps after the laser pulse. The excellent overlap of the properly scaled FTIR

spectrum of ( 1) with the transient spectrum observed within 40 – 3000 ps convinces us

that no absorptions of ketene product have been detected in this region. The broad

positive band observed in the first 40 ps post laser pulse is assigned to the hot ground

state. We conclude that 31% of the diazo precursor is recovered with 270 nm light

excitation. The recovery time constant monitored at 2083 cm -1 is 18.4 ps ( Figure 7.13 ),

which is due to vibrational cooling.

Figure 7.12. Transient IR spectra produced by ultrafast time-resolved LFP ( λex = 270 -1 nm) of PhCHCHCN 2CO 2CH 3 in chloroform with a spectral window of 2160 – 2000 cm . The dotted curve is scaled conventional FTIR spectra in chloroform.

194

Figure 7.13. The kinetic trace of the diazo band obtained by ultrafast time-resolved LFP (λex = 270 nm) of PhCHCHCN 2CO 2CH 3 in chloroform. The kinetic trace is monitored at 2095 cm -1 and is fitted to an exponential function.

Figure 7.14. Transient IR spectra produced by ultrafast time-resolved LFP ( λex = 270 nm) of PhCHCHCN 2CO 2CH 3 ( 1) in chloroform with a spectral window of 1750 – 1640 cm -1. The dotted curve is the stationary FTIR spectrum in chloroform.

The transient IR spectra shown in Figure 7.14 were obtained by focusing the

spectral window in the carbonyl vibration region (1750 – 1640 cm -1). The negative band

195 detected in the center of the spectral window is due to the bleaching of the carbonyl

(C=O) stretch of precursor ( 1). The decay of the positive band (~ 1650 cm -1) is observed to the red of the negative band, and the recovery of the bleaching band, immediately after laser pulse (300 fs) resembles the behavior of hot ground state recovery observed in the diazo stretching region. However, some important differences are noticed. An isosbestic point at 1680 cm -1 is seen as the hot ground state decays and the bleaching band recovers

over 12 ps, whereas no isosbestic point was observed previously in the diazo stretching

region. The carbonyl bleaching band reshapes over 65 ps, clearly indicating that a new

species absorbing around 1705 cm -1 is formed rapidly. The evolution of the 1705 cm -1 band is complete 65 ps after the laser flash, and retains its intensity over 3 ns. We posit that the most probable candidate for this band is 3-phenyl-1-carbomethoxy cyclopropene

(3) formed by cyclization of the singlet carbene 12. Calculations predict cyclopropene ( 3)

has an intense band at 1723 cm -1 ( Table 7.1), in excellent agreement with the observed band at 1705 cm -1 (Figure 7.14 ).

In addition, calculations also predict cyclopropene ( 3) has another intense

absorption band at 1776 cm -1. Both the 1723 and 1776 cm -1 bands are due to the coupled

C=O and C=C vibrational modes in cyclopropene ( 3) ( Table 7.1). Based on these computational results, ultrafast LFP studies of PhCH=CHCN 2CO 2CH 3 (1) in chloroform

-1 (λex = 270 nm) were performed in the spectral window of 1790 – 1680 cm , and the transient IR spectra produced shown in Figure 7.15 . Just as predicted, a positive band at

1770 cm -1 was observed. This band narrows and shifts to the blue within 80 ps of the laser pulse and retains the same intensity over 3 ns. Identical kinetic behavior monitored at 1770 and 1705 cm -1 suggests that both bands are associated with the same intermediate.

196 The direct observation of both vibrational bands and their excellent agreement with calculations supports the assignment of transient spectra to cyclopropene product ( 3).

197

λmax f Cartesian Coordinates C -3.696350000 0.154856000 0.751664000 327.3 0.0251 C -2.531298000 0.921345000 0.715043000 C -1.425086000 0.515355000 -0.042721000 C -1.514867000 -0.682470000 -0.766484000 281.6 0.0064 C -2.678035000 -1.450281000 -0.732064000 C -3.774605000 -1.035270000 0.026939000 267.6 0.0014 H -4.542312000 0.487731000 1.347607000 H -2.475358000 1.847711000 1.282966000 239.1 0.0171 H -0.667230000 -1.017026000 -1.359092000 H -2.727895000 -2.376626000 -1.298811000 H -4.680486000 -1.634876000 0.053974000 236.4 0.1987 C -0.192603000 1.370295000 -0.063032000 C 0.714145000 1.434952000 -1.274091000 217.8 0.1262 H 0.809841000 1.754482000 -2.302065000 C 1.190113000 0.795064000 -0.248006000 209.2 0.0002 C 2.251932000 -0.000900000 0.376650000 O 2.200182000 -0.466315000 1.495146000 207.7 0.0003 O 3.304739000 -0.139552000 -0.460948000 C 4.405014000 -0.896859000 0.070213000 H 5.156283000 -0.911390000 -0.719716000 199.6 0.0434 H 4.801507000 -0.419003000 0.970216000 H 4.087506000 -1.913073000 0.319036000 199.1 0.0011 H -0.291381000 2.272474000 0.547632000 Frequency IR intensity Frequency IR intensity Frequency IR intensity (cm -1) (km/mol) (cm -1) (km/mol) (cm -1) (km/mol) 26 0.8 784 11.1 1317 0.5 31 0.6 795 11.9 1360 0.6 47 0.3 828 0.0 1433 6.8 122 2.0 871 17.9 1447 7.9 136 1.4 892 0.2 1454 6.2 144 2.2 929 0.0 1465 10.0 168 0.8 949 2.9 1487 12.4 232 1.9 955 0.1 1581 1.0 286 11.3 977 0.8 1601 8.4 324 11.4 1012 19.8 1723 176.5 354 1.2 1020 9.8 1776 187.4 402 0.0 1033 12.3 2956 34.1 453 2.1 1070 5.8 2967 38.1 529 5.9 1119 124.7 3027 19.8 552 1.8 1137 1.0 3051 8.7 598 2.5 1146 0.3 3057 0.6 612 0.5 1167 1.2 3061 17.8 677 59.7 1171 2.0 3067 11.0 687 20.1 1197 254.1 3074 42.3 746 17.2 1237 272.3 3085 24.5 768 4.2 1275 3.4 3161 0.4

Table 7.1. Geometry optimization and frequency analyses for cyclopropene product ( 3) were obtained with the B3LYP/6-31G(d) level of theory. Frequencies were scaled by a factor of 0.9614. Vertical transition energies (nm) and oscillator strength ( f) were calculated at the TD-B3LYP/6-311+G(d,p)//B3LYP/6-31G(d) level of theory.

198

Figure 7.15. Transient IR spectra produced by ultrafast time-resolved LFP ( λex = 270 nm) of PhCHCHCN 2CO 2CH 3 (1) in chloroform with a spectral window of 1790 – 1680 cm -1. The dotted curve is the stationary FTIR spectra in chloroform.

A structurally similar cyclopropene 1-carbomethoxy-3,3-dimethyl-2-phenyl

cyclopropene 45 was synthesized with IR bands (neat) at 1840 and 1710 cm -1. This is consistent with the transient bands observed for cyclopropene ( 3) in this study.

Subtraction of the scaled FTIR spectrum of the diazo precursor from the transient spectra observed above produces the spectra shown in Figure 7.16 . Clearly two positive bands at

1770 cm -1 and 1705 cm -1 are evident and persist for at least 3 ns. FTIR spectrum of a chloroform solution of ( 1) indicates that a new band at 1770 cm -1 is present after laser

flash photolysis experiments ( Figure 7.17 ), demonstrating that the carriers of these bands are persistent species

199

Figure 7.16. Transient spectra at selected time delays obtained by subtracting the bleaching band of precursor from the transient spectra produced by ultrafast photolysis (270 nm) of PhCHCHCN 2CO 2CH 3 in chloroform with a spectral window of 1790 – 1680 cm -1.

Figure 7.17. Stationary FTIR spectra of PhCHCHCN 2CO 2CH 3 (1) in chloroform taken before and after laser flash photolysis (270 nm).

Previously we have shown that integration of band intensities can provide precise kinetic information regarding the formation of newly born species undergoing vibrational cooling. 34,46 Integration of both bands of cyclopropene in the spectral range 1790 – 1680 cm -1 ( Figure 7.18 ) indicates that cyclopropene ( 3) is formed with a time constant of 32 ±

8 ps in chloroform ( Table 7.4,), which is in excellent agreement with the decay lifetime

200 (τ = 33.3 ± 2.5 ps) of singlet carbene ( 12) observed in this solvent obtained by ultrafast time-resolved UV–vis spectroscopy ( Figure 7.4c). The same time constant obtained with two different time-resolved techniques with two different excitation wavelength confirms that singlet carbene ( 12) is the immediate precursor of cyclopropene ( 3).

Figure 7.18. Band intensity integration of the transient spectra of cyclopropene ( 1) obtained from ultrafast LFP ( λex = 270 nm) of PhCHCHCN 2CO 2CH 3 in chloroform with a spectral window of 1790 – 1680 cm -1.

B3LYP/6-31G(d) calculations predict the C=O stretching mode of singlet carbene

(12) at 1641 cm -1 ( Table 7.2) and the triplet ( 32) at 1653 cm -1 (Table 7.2). Both singlet and triplet carbenes have been previously observed by ultrafast LFP with UV-vis detection. However, in the 1650 – 1640 cm -1 spectral range where the carbenes are predicted (Figure 7.14 ), only the broad spectra of hot ground state is observed after the laser pulse. But we can not rule out some contribution from singlet carbenes because of the evolution of vibrational cooling in this spectral region. Therefore more studies are needed in this spectral region to detect carbenes.

201

λmax f Cartesian Coordinates C -3.564923000 1.463671000 -0.102973000 899.7 0.0416 C -2.186586000 1.292269000 -0.043356000 C -1.614840000 0.000458000 -0.010898000 C -2.482766000 -1.115359000 -0.004945000 377.6 0.1220 C -3.859441000 -0.939895000 -0.062643000 C -4.405989000 0.347675000 -0.114855000 337.6 0.0057 H -3.985673000 2.464631000 -0.134005000 H -1.530523000 2.158958000 -0.033668000 327.6 0.0186 H -2.072837000 -2.117656000 0.064673000 H -4.513851000 -1.806982000 -0.055437000 H -5.483827000 0.478832000 -0.152365000 317.5 0.0008 C -0.173862000 -0.113540000 -0.032539000 H 0.375428000 0.815914000 0.107225000 310.3 0.4164 C 0.573763000 -1.278615000 -0.092174000 H 0.064693000 -2.230868000 -0.253571000 297.3 0.2051 C 1.890616000 -1.294259000 0.354760000 C 2.790557000 -0.176050000 0.343434000 248.0 0.0075 O 2.793514000 0.679508000 1.224208000 O 3.700657000 -0.235608000 -0.660788000 C 4.733860000 0.760556000 -0.606184000 239.4 0.0119 H 5.388685000 0.544557000 -1.451249000 H 4.313745000 1.766254000 -0.698536000 235.4 0.0126 H 5.288456000 0.693695000 0.333818000 Frequency IR intensity Frequency IR intensity Frequency IR intensity (cm -1) (km/mol) (cm -1) (km/mol) (cm -1) (km/mol) 38 1.2 819 0.5 1323 10.8 43 3.8 820 54.7 1389 188.9 59 0.1 827 4.6 1431 38.4 95 4.9 876 13.2 1442 47.4 106 0.8 903 8.0 1454 4.5 121 2.9 937 3.2 1469 7.7 149 4.9 954 36.8 1482 3.3 183 3.1 968 6.9 1498 57.9 255 18.5 975 8.4 1565 2.2 299 9.9 996 16.0 1591 12.4 386 20.6 1015 0.1 1641 181.4 398 1.6 1074 4.8 2952 52.5 407 47.5 1135 6.4 3006 16.6 485 81.9 1150 1.2 3021 22.8 505 137.9 1159 427.0 3054 23.3 573 44.0 1167 144.3 3060 3.3 606 0.0 1173 429.9 3066 4.6 614 5.5 1200 152.4 3072 0.2 670 23.7 1263 4.0 3081 13.7 733 43.4 1287 10.6 3090 25.1 748 24.0 1314 77.8 3097 15.1

Table 7.2. Geometry optimization and frequency analyses for singlet carbene 1 1 PhCH=CHCCO 2CH 3 ( 2) were obtained with the B3LYP/6-31G(d) level of theory. Frequencies were scaled by a factor of 0.9614. Vertical transition energies (nm) and oscillator strength ( f) were calculated at the TD-B3LYP/6-311+G(d,p)//B3LYP/6-31G(d) level of theory.

202

λmax f Cartesian Coordinates C 3.804939000 1.352851000 -0.000153000 457.3 0.0001 C 2.416872000 1.280410000 -0.000151000 C 1.746017000 0.034051000 0.000033000 C 2.539338000 -1.138574000 0.000184000 391.7 0.0028 C 3.926116000 -1.061807000 0.000171000 C 4.568705000 0.181992000 0.000008000 359.5 0.0009 H 4.294592000 2.322947000 -0.000285000 H 1.825570000 2.192698000 -0.000278000 358.7 0.0002 H 2.064088000 -2.114832000 0.000289000 H 4.514216000 -1.975699000 0.000276000 H 5.653767000 0.235973000 -0.000006000 341.8 0.5761 C 0.302841000 0.029415000 0.000025000 H -0.193024000 0.995831000 0.000355000 318.8 0.1518 C -0.524596000 -1.100491000 -0.000248000 H -0.060202000 -2.087694000 -0.000470000 313.7 0.0183 C -1.893990000 -1.079551000 -0.000148000 C -2.898380000 -0.038510000 0.000028000 305.8 0.0001 O -2.646659000 1.159936000 0.000278000 O -4.154059000 -0.537556000 -0.000121000 C -5.201766000 0.443939000 0.000030000 301.4 0.0598 H -6.132263000 -0.124437000 -0.000136000 H -5.138834000 1.076718000 0.889846000 292.7 0.0000 H -5.138743000 1.077092000 -0.889513000 Frequency IR intensity Frequency IR intensity Frequency IR intensity (cm -1) (km/mol) (cm -1) (km/mol) (cm -1) (km/mol) 36 1.8 808 0.3 1318 0.8 41 0.2 826 2.7 1403 146.6 64 0.6 835 0.9 1427 7.3 88 0.3 861 9.7 1446 157.6 127 0.5 877 7.2 1454 5.9 155 3.0 895 28.7 1457 7.4 158 1.1 928 0.3 1466 4.1 220 4.2 956 0.4 1483 13.0 252 0.6 969 0.4 1557 2.1 323 1.0 991 15.6 1579 5.5 324 30.9 1014 0.3 1653 103.8 395 0.0 1070 5.2 2953 43.4 397 2.0 1137 0.8 3018 5.7 486 3.6 1146 1.2 3023 23.2 502 3.0 1157 84.9 3058 23.9 606 0.1 1165 23.7 3060 4.5 613 1.3 1175 19.2 3066 0.3 668 20.9 1200 530.6 3075 11.4 678 5.4 1219 179.4 3083 31.1 708 24.5 1260 149.9 3086 5.2 738 24.0 1302 104.6 3090 22.5

Table 7.3. Geometry optimization and frequency analyses for triplet carbene 3 3 PhCH=CHCCO 2CH 3 ( 2) were obtained with the B3LYP/6-31G(d) level of theory. Frequencies were scaled by a factor of 0.9614. Vertical transition energies (nm) and oscillator strength ( f) were calculated at the TD-B3LYP/6-311+G(d,p)//B3LYP/6-31G(d) level of theory.

203

Figure 7.19. Transient IR spectra produced by ultrafast time-resolved LFP ( λex = 270 nm) of PhCHCHCN 2CO 2CH 3 (1) in chloroform with a spectral window of 1800 – 1560 cm -1. The transient spectra were obtained by combining the data of multiple experiments with different spectral windows. The dotted curve is the stationary FTIR spectra in chloroform. The green bars are vibrational frequencies of cyclopropene (3) and singlet (12) and triplet ( 32) carbenes calculated at the B3LYP/6-31G(d) level of theory and scaled using the experimental (FTIR) frequencies of diazo precursor (1). The calculated intensities were scaled.

The transient spectra shown in Figure 7.19 are produced by combining the data of multiple experiments in the spectral window 1800 – 1560 cm -1. As discussed above, the

two bands at 1770 and 1705 cm -1 are due to cyclopropene (3). There is a noticeable

correlation between the decay of the broad band (1600 – 1680 cm -1) and the growth of the cyclopropene bands (1790 – 1670 cm -1), and an isosbestic point at 1684 cm -1 is observed, suggesting that the singlet carbene is probably detected in the 1600 – 1680 cm -1 spectral

region. In addition, the absorption at ~ 1650 cm -1 is present 3 ns after the laser pulse,

consistent with the triplet carbene (32) detected in the UV-vis region. However, precise kinetics of either the singlet carbene ( 12) or triplet carbene (32) can not be obtained by time-resolved IR methods as they are greatly affected by the overlapping hot ground state

204 (1#). The long lifetime of the triplet, relative to the singlet carbene indicates that the two

spin isomers are not in rapid equilibrium.

7.2.4. Ultrafast Time-resolved IR Studies in Methanol-OD

In methanol-OD ultrafast LFP ( λex = 270 nm) of PhCH=CHCN 2CO 2CH 3 (1)

produces the transient spectra (1750 – 1630 cm -1) shown in Figure 7.20 . As in

chloroform, the negative beaching band of the C=O stretching band overlapped severely

with the cyclopropene ( 3) band; the hot ground state band on the red edge and an isosbestic point at 1684 cm -1 are observed. In methanol-OD the cyclopropene band is split into a doublet. The splitting of the C=O stretch of esters by alcohols was reported by

Minato, 47 and is consistent with our recent studies of lactams and amides. 43 Significantly, in the 1680 – 1580 cm -1 spectral range the transient absorption of singlet carbene ( 12) is more prominent, and that its decay correlates very well with the growth of the band of cyclopropene in the bleaching region ( Figure 7.21 ). Fitting of the kinetic traces at 1710

-1 and 1648 cm produces the same biexponential function with time constants of τ1 = 7.0

± 0.1 ps and τ2 = 95 ± 12 ps, which is in excellent agreement with the singlet carbene lifetime ( Figure 7.10b, τ = 91 ± 4.4 ps) detected in the UV-vis spectral region in the same solvent ( Figure 7.9). The fast component is assigned to vibrational cooling. The stronger transient absorption of singlet carbene obtained in methanol-OD relative to chloroform is in good agreement with the extended singlet carbene lifetime and lower triplet carbene yield in methanol-OD.

205

Figure 7.20. Transient IR spectra produced by ultrafast time-resolved LFP ( λex = 270 nm) of PhCHCHCN 2CO 2CH 3 (1) in methanol-OD with a spectral window of 1750 – 1630 cm -1.

Figure 7.21. Kinetic traces monitored at 1684, 1648, and 1710 cm-1 of transient IR spectra obtained by ultrafast LFP ( λex = 270 nm) of PhCHCHCN 2CO 2CH 3 (1) in methanol-OD. Fitting of the kinetic traces at 1648 and 1710 cm-1 produces the same time -1 constants of τ1 = 7.0 ± 0.1 ps and τ2 = 95 ± 12 ps. The flat trace at 1648 cm indicates the isosbestic point.

206 Ultrafast UV-vis Ultrafast IR Solvents Singlet carbene ( 12) Cyclopropene ( 3) ACN 25.7 ± 1.6 – MeOH 38.4 ± 2.4 – MeOD 91.1 ± 4.4 95 ± 12 CHCl 3 33.3 ± 2.5 32 ± 8 CHX 6.5 ± 1.1 –

Table 7.4. Summary of lifetimes ( τ /ps) of transients obtained by ultrafast LFP of PhCH=CHCN 2CO 2CH 3 (1) with UV-Vis detection ( λex = 310 nm) and IR detection ( λex = 270 nm) in acetonitrile (ACN), methanol, methanol-OD, chloroform, and cyclohexane (CHX).

The experimental data obtained from ultrafast LFP with UV-vis and IR studies are tabulated in Table 7.4. The singlet and triplet carbenes were detected in the UV-vis

region, and the cyclopropene product was detected in the infrared region. The decay of

singlet carbene (12) is correlated with the formation of cyclopropene (3) through the use

of two different techniques.

7.3. Theoretical Studies

DFT calculations with B3LYP/6-311+G(d,p) predict the singlet-triplet energy gap

3 ∆EST is 7.9 kcal/mol with triplet ( 2) being the ground state. As a reference, the same level of theory predicts ∆EST = 12.0 kcal/mol for cis -vinylmethylene (11.1 kcal/mol predicted for trans ), consistent with previous results. 21-25 In contrast, vinylchlorocarbene

26 is a ground-state singlet species with ∆EST = -2.1 kcal/mol. The large ∆EST predicted for styrylcarbomethoxy carbene ( 2) indicates that the two spin isomers are not in

equilibrium, and that intersystem crossing is unidirectional. This result is in contrast with

35,48 the observation of 2-naphthylcarbomethoxy carbene (2-NpCCO 2CH 3), in which the

49,50 ∆EST is near zero and ground-state carbene is dependent on solvent polarity. This

indicates that the vinyl group is less effective than the aryl group in stabilizing singlet

207 carbene relative to its triplet state. This is understandable since the singlet carbene is a zwitterionic species which is stabilized by aryl groups, while the triplet carbene is a biradical-like species and is effectively stabilized through delocalization by vinyl groups.

Singlet Triplet Singlet Triplet vinylmethylene vinylmethylene vinylchlorocarbene vinylchlorocarbene

Singlet styrylcarbomethoxycarbene Triplet styrylcarbomethoxycarbene

Figure 7.22. Selected structural parameters calculated for singlet and triplet states of vinylmethylene, vinylchlorocarbene, and styrylcarbomethoxycarbene ( 2) at the B3LYP/6-31G(d) level of theory in the gas phase. Bond lengths are in angstroms. Bond angles are in degrees.

Triplet vinylcarbene has been predicted to exist as a highly delocalized allylic diradical. 51-53 In this work the singlet and triplet vinylcarbenes 2 were computed at the

UB3LYP/6-31G(d) level of theory and the predicted geometry parameters are shown in

Figure 7.22 . Vinylmethylene and vinylchlorocarbene were also computed for the sake of

comparision. Both the triplet states of vinylmethylene and vinylchlorocarbene are highly

delocalized, as evident from indistinguishable C=C and C–C bond lengths (1.37 – 1.39

208 Å). In comparison with the partially delocalized singlet vinylmethylene, singlet vinylchlorocarbene is highly localized, as shown from the typical short C=C bond length

(1.35 Å) and long C–C bond length (1.45 Å). 26 However, a major delocalization is predicted in singlet vinyl carbene 12, with the same bond length (1.39 Å) for C=C and C–

C bonds. This indicates that the zwitterionic singlet vinylcarbene is delocalized over the conjugated C=C double bond. Thus the extent of delocalization in singlet vinylchlorocarbene and vinylcarbene 12 is significantly different, which accounts for the difference in reactivity. The reversed order in the bond lengths of the C=C and C–C bonds of triplet vinylcarbene 32 reflects the fact that the free radical character, prefers to

localize on the benzylic carbon position. The triplet vinylcarbene 32 geometry is planar, while the carbene’s ester group is still near orthogonal, as other carbene esters reported. 54

The lifetime of singlet vinylcarbene ( 12) is considerably shorter than those of

1 55 previously studied singlet carbenes NpCCO 2CH 3 ( τ = 830 ns in Freon-113),

1 37 1 34 BpCCO 2CH 3 ( τ =180 ns in acetonitrile), and PhCH ( τ =414 ps in chloroform). The significantly shortened lifetime of singlet vinylcarbene ( 12) reflects its higher reactivity.

In addition to intersystem crossing to the lower energy triplet state, the singlet vinylcarbene ( 12) has two other intramolecular deactivation pathways in solution: cyclization to cyclopropene and Wolff rearrangement to ketene. These two reactions were studied computationally in the gas phase ( Figure 7.23 ). The transition state predicted for

cyclization is only 3.1 kcal/mol above the singlet vinylcarbene ( 12); however, the activation energy for ketene formation is predicted to be 11.3 kcal/mol. Thus cyclopropene formation is more rapid than the Wolff rearrangement, from the singlet carbene. The Arrhenius activation energies ( Ea) have been experimentally determined for

209 cyclization in (1-methylvinyl)chlorocarbene ( Ea = 6.6 kcal/mol) and (1-

chlorovinyl)chlorocarbene ( Ea = 5.7 kcal/mol) by Moss et al. using the pyridine-ylide method. 26,27 The smaller activation energy predicted for singlet vinylcarbene ( 12) is consistent with its short lifetimes as measured in this study. This also confirms that stabilization of singlet vinylchlorocarbenes by halogen substitution increases the energy barrier to cyclization.

Figure 7.23. The reaction paths predicted for the formation of cyclopropene (TS1) and 1 ketene (TS2) from singlet vinylcarbene PhCHCHCCO 2CH 3. Relative free energies ∆G (298 K) were calculated at the B3LYP/6-31G(d) level of theory in the gas phase.

The solvent effect on the reactions of singlet carbene was studied with the PCM model. Calculations predict that the activation energies for cyclization depend on solvent polarity. The activation energy in cyclohexane (3.4 kcal/mol) is close to that in the gas phase (3.0 kcal/mol), however, it increases to 4.9 kcal/mol in methanol and to 4.8 kcal/mol in acetonitrile. Transition state theory predicts a lifetime of 30 ps for the singlet carbene in the gas phase (298 K). This is in fair agreement with the measured lifetime of

12 ( τ = 6.5 ± 1.1 ps) in cyclohexane ( Table 7.4).

210 a b b b b b Gas phase Gas phase Methanol CHCl 3 ACN CHX TS1 3.1 3.0 4.9 0.6 4.8 3.4 Singlet carbene 0.0 0.0 0.0 0.0 0.0 0.0 Triplet carbene -8.6 -7.9 -3.0 -4.8 -3.0 -6.4 TS2 11.3 12.1 14.4 13.4 14.3 12.6 cyclopropene -21.7 -22.3 -18.2 -19.8 -18.3 -21.1 ketene -27.2 -27.4 -21.5 -23.6 -21.6 -25.6 a Calculated Gibbs free energies ∆G (298 K) with ZPE included.

Table 7.5. Relative energies ∆E (0 K) calculated for singlet vinylcarbene ( 12), triplet vinylcarbene ( 32), the transition states for the formation of cyclopropene (TS1) and ketene (TS2) from singlet vinylcarbene ( 12) and their related products. The gas phase calculations were performed at the B3LYP/6-31G(d)// B3LYP/6-31G(d) level of theory. The PCM model is used in computing the singlet point energies in methanol, chloroform (CHCl 3), acetonitrile (ACN), and cyclohexane (CHX) with geometries optimized in the gas phase. aZero point energies were not included. bZero point energies were included. The energies (kcal/mol) are relative to singlet carbene (12).

211

Energy = -575.512797 C 3.454494000 -1.412157000 -0.253869000 C 2.129066000 -1.191161000 -0.607965000 C 1.510124000 0.053940000 -0.355106000 C 2.260700000 1.056992000 0.298722000 C 3.585070000 0.829474000 0.653617000 C 4.188087000 -0.402034000 0.376496000 H 3.916827000 -2.373333000 -0.460374000 H 1.556934000 -1.976349000 -1.095924000 H 1.788589000 2.003041000 0.545262000 H 4.149386000 1.608537000 1.158632000 H 5.221572000 -0.577617000 0.661857000 C 0.162494000 0.276744000 -0.819716000 C -0.568767000 1.491239000 -0.692014000 H -0.167863000 2.499853000 -0.800726000 C -1.626921000 1.126509000 0.029511000 C -2.529383000 0.071926000 0.367895000 O -2.377964000 -0.734543000 1.274087000 O -3.658690000 0.151925000 -0.396727000 C -4.709307000 -0.743771000 -0.006903000 H -5.543420000 -0.519066000 -0.673116000 H -4.398282000 -1.785912000 -0.124241000 H -4.996415000 -0.578030000 1.035426000 H -0.368045000 -0.586381000 -1.209680000 Frequency IR intensity Frequency IR intensity Frequency IR intensity (cm -1) (km/mol) (cm -1) (km/mol) (cm -1) (km/mol) -123 40.1 818 9.0 1374 3.1 42 2.2 833 23.7 1454 28.9 53 1.4 852 39.9 1487 7.0 76 7.4 855 8.6 1511 12.4 109 2.3 884 21.4 1511 2.3 124 0.4 932 3.5 1527 14.5 140 4.7 972 0.0 1540 1.9 151 0.8 1002 0.0 1626 1.1 240 2.8 1013 1.0 1651 0.2 312 11.3 1038 13.2 1687 109.7 396 9.1 1054 0.4 1757 435.3 415 0.1 1116 2.6 3069 55.7 430 26.2 1181 9.4 3131 32.0 549 22.8 1193 321.7 3140 22.7 559 93.4 1195 16.2 3174 25.4 567 0.7 1207 568.1 3186 2.9 631 1.8 1213 57.6 3192 0.1 631 3.4 1220 268.4 3202 11.6 699 17.9 1275 5.3 3209 15.3 710 57.1 1296 109.1 3211 24.9 769 24.5 1358 23.8 3218 15.5

Table 7.6. Transition state (TS1) for the formation of cyclopropene ( 3) from the singlet carbene optimized at the B3LYP/6-31G(d) level of theory. Frequencies are not scaled.

212

Energy = -575.4981469 C 3.654729000 1.201453000 0.458462000 C 2.274491000 1.256588000 0.287199000 C 1.546299000 0.120446000 -0.119719000 C 2.257759000 -1.070141000 -0.375684000 C 3.636421000 -1.123836000 -0.202514000 C 4.340855000 0.009438000 0.216953000 H 4.195786000 2.088268000 0.776520000 H 1.742066000 2.185468000 0.477185000 H 1.729690000 -1.950324000 -0.729156000 H 4.168170000 -2.049229000 -0.406048000 H 5.418936000 -0.035274000 0.343773000 C 0.097369000 0.224927000 -0.230712000 H -0.307231000 1.234036000 -0.162943000 C -0.788007000 -0.791167000 -0.457610000 H -0.438728000 -1.822713000 -0.496352000 C -2.117358000 -0.534451000 -0.853138000 C -3.069634000 0.404819000 -0.656093000 O -3.674248000 1.409193000 -0.878569000 O -3.539794000 -0.745933000 0.319487000 C -3.325649000 -0.376001000 1.691327000 H -3.742352000 -1.181528000 2.301204000 H -2.255159000 -0.271728000 1.909018000 H -3.842485000 0.564501000 1.908621000 Frequency IR intensity Frequency IR intensity Frequency IR intensity (cm -1) (km/mol) (cm -1) (km/mol) (cm -1) (km/mol) -334 90.5 772 43.0 1373 2.0 38 1.6 852 0.3 1437 39.8 49 2.9 859 12.5 1472 12.9 60 0.2 889 16.8 1496 21.2 104 8.4 929 33.6 1516 4.8 110 14.0 934 5.8 1521 1.4 139 4.9 970 2.2 1544 12.4 189 6.9 994 75.0 1605 71.9 247 35.2 1002 129.7 1636 9.6 275 4.9 1007 7.2 1659 0.4 335 3.0 1015 2.9 2048 854.1 414 4.2 1058 0.6 3040 39.7 420 14.0 1116 3.5 3117 16.4 431 12.3 1175 1.1 3150 12.8 520 17.0 1183 3.9 3151 13.0 548 3.4 1195 0.1 3170 12.3 614 22.3 1216 4.3 3184 6.6 633 0.6 1240 9.0 3191 0.2 639 152.5 1316 0.6 3201 12.5 699 64.7 1341 4.9 3210 27.4 702 13.4 1368 3.5 3217 20.6

Table 7.7. Transition state (TS2) for the formation of ketene from the singlet carbene optimized at the B3LYP/6-31G(d) level of theory. Frequencies are not scaled.

213 7.4. Conclusions

In summary, we have observed a singlet vinyl carbene ( 12) produced directly from

the excited state of a diazo precursor ( 1). The singlet carbene undergoes intramolecular cyclopropenation reactions to produce the cyclopropene product ( 3) in acetonitrile, chloroform and cyclohexane, and undergoes intersystem crossing to the ground state triplet carbene ( 32) to a small extent. The triplet carbene is not in rapid equilibrium with

the singlet due to a relatively large singlet triplet separation. Calculations predict that

both singlet and triplet vinylcarbene 2 are highly delocalized.

7.5. Calculations

DFT and TD-DFT calculations were performed using the Gaussian03 suite of programs 32 at The Ohio Supercomputer Center. Geometries were optimized at the

B3LYP/6-31+G(d) level of theory with single-point energies obtained at the B3LYP/6-

311+G(d,p)//B3LYP/6-31+G(d) level of theory. Vibrational frequency analyses at the

B3LYP/6-31+G(d) level were used to verify that stationary points obtained corresponded

to energy minima, and were scaled with a factor of 0.9614. 56 The electronic spectra were computed using TD-DFT with Gaussian03 at the B3LYP/6-311+G(d,p) level and using the B3LYP/6-31+G(d) geometry; for these calculations, 10 allowed electronic transitions were calculated. The solution phase calculations used PCM models and were performed at the B3LYP/6-31G(d)(PCM)//B3LYP/6-31G(d)(PCM) level of theory. Transition states were localized with QST2 methodology and confirmed by the presence of only one imaginary frequency. Intrinsic reaction coordinate (IRC) calculations 57,58 were performed

to confirm the reactant and product associated with these transition states.

214 7.6. Ultrafast Spectroscopy

Ultrafast IR spectroscopic studies were performed using the home-built pump- probe spectrometer described previously.37,59 Samples were prepared in 50 mL of solvent

with unit absorption at the excitation wavelength with 1.0 mm optical length. The entire

set of pump-probe delay positions (cycle) is repeated at least three times, to achieve good

data reproducibility from cycle to cycle. To avoid rotational diffusion effects, the angle

between polarizations of the pump beam and the probe beam was set to the magic angle

(54.7°). Kinetic traces are analyzed by fitting to exponential terms. All experiments were

performed at room temperature.

7.7. Synthesis

All materials and solvents were purchased from Sigma-Aldrich, Inc. The solvents

for ultrafast studies were spectrophotometric grade from Sigma-Aldrich, Inc. and used as

received.

CH N DBU, TsN N2 OH 2 2 OCH3 3 OCH3 O O O

60 A. Diazomethane (CH 2N2). To 20 ml of anhydrous ether is added 6 mL of 40%

KOH aqueous solution and the mixture is cooled to 5 ºC with ice-water. With continuous

cooling and stirring, 2 g of finely powered nitrosomethylurea in small portions is added

over one minute and subsequent portions are not added until the previous portion has

dissolved. After continual stirring for 2 min, the yellow ether layer containing about 0.5 g

(12 mmol) of diazomethane is decanted and dried over KOH pellets in freezer for one

hour.

215 B. Methyl trans -styrylacetate. Trans -styrylacetic acid (1.62 g, 10 mmol) was

dissolved in 20 mL of anhydrous ether and cooled with ice-water bath. Then the ethereal

solution of diazomethane prepared above was added slowly by a pipette. Bubbles come

out immediately upon addition. The addition continued until no bubbles were visible,

stirring is then continued for 1 hr. Removal of the solvent with a rotary evaporator yields

a yellow oil. GC-MS: 176 (M +, 100%), 134 (60%), 144 (20%). The ester is used directly

1 in the next step without further purification. H-NMR (400 MHz, CDCl 3): δ 3.29 (dd, 2H,

J1 = 1.2 Hz, J2 = 7.2 Hz), 3.75 (s, 3H), 6.33 (dt, 1H, J1 = 7.2 Hz, J2 = 16.0 Hz), 6.53 (d,

1H, J = 15.6 Hz), 7.24-7.28 (m, 1H), 7.32-7.36 (m, 2H), 7.40-7.42 (m, 2H). 13 C-NMR

(100 MHz, CDCl 3): δ 172.01, 136.82, 133.50, 128.55, 127.58, 126.30, 121.67, 51.94,

38.24.

C. Methyl ( E)-2-Diazo-4-phenylbutenoate .31 DBU (1.64 mL, 11 mol) was

added dropwise via a syringe to a stirred solution of methyl trans -styrylacetate prepared

above (10 mmol) and p-acetamidobenzenesulfonyl azide (ABSA, 1.67 g, 11 mmol) in

acetonitrile (20 mL) at 0 ºC over 30 min. After stirring the mixture for 4 hr, saturated

NH 4Cl solution was added and the mixture was extracted twice with CH 2Cl 2. The organic layer was dried with Na 2SO 4 and the solvent was evaporated under reduced pressure at 25

ºC. The residue was purified with chromatography and 1.2 g of a red oil was obtained

-1 1 with column eluant hexane /ethyl acetate (15:1). IR (CHCl 3) 2095, 1700, 1630 cm . H-

NMR (400 MHz, CDCl 3): δ 3.88 (s, 3H), 6.23 (d, 1H, J = 16.4 Hz), 6.51 (d, 1H, J = 16.0

13 Hz), 7.21-7.25 (m, 1H), 7.32 (m, 4H). C-NMR (100 MHz, CDCl 3): δ 165.59, 136.82,

128.71, 127.11, 125.88, 123.11, 111.32, 52.32.

216 7.8. References for Chapter 7

1. Moss, R. A.; Platz, M. S.; Jones, M., Jr. Reactive Intermediate Chemistry ; Wiley: New York, 2004.

2. Platz, M. S. Kinetics and Spectroscopy of Carbenes and Biradicals ; Plenum: New York, 1990.

3. Jones, M., Jr.; Rettig, K. R. J. Am. Chem. Soc. 1965 , 87 , 4015-4016.

4. Closs, G. L.; Closs, L. E. J. Am. Chem. Soc. 1961 , 83 , 2015-2016.

5. Zupancic, J. J.; Schuster, G. B. J. Am. Chem. Soc. 1980 , 102 , 5958-5960.

6. Wang, J.; Kubicki, J.; Hilinski, E. F.; Mecklenburg, S. L.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2007 , 129 , 13683-13690.

7. Joines, R. C.; Turner, A. B.; Jones, W. M. J. Am. Chem. Soc 1969 , 91 , 7754-7755.

8. West, P. R.; Chapman, O. L.; LeRoux, J. P. J. Am. Chem. Soc. 1982 , 104 , 1779- 1782.

9. McMahon, R. J.; Abelt, C. J.; Chapman, O. L.; Johnson, J. W.; Kreil, C. L.; LeRoux, J. P.; Mooring, A. M.; West, P. R. J. Am. Chem. Soc. 1987 , 109 , 2456- 2469.

10. McMahon, R. J.; Chapman, O. L. J. Am. Chem. Soc. 1987 , 109 , 683-692.

11. Chapman, O. L.; McMahon, R. J.; West, P. R. J. Am. Chem. Soc. 1984 , 106 , 7973-7974.

12. Warmuth, R.; Marvel, M. A. Chemistry--A European Journal 2001 , 7, 1209-1220.

13. Schreiner, P. R.; Karney, W. L.; Schleyer, P. v. R.; Borden, W. T.; Hamilton, T. P.; Schaefer, H. F., III. J. Org. Chem. 1996 , 61 , 7030-7039.

14. Hutton, R. S.; Manion, M. L.; Roth, H. D.; Wasserman, E. J. Amer. Chem. Soc. 1974 , 96 , 4680-4682.

15. Senthilnathan, V. P.; Platz, M. S. J. Am. Chem. Soc. 1981 , 103 , 5503-5511.

16. Cochran, E. L.; Adrian, F. J.; Bowers, V. A. J. Chem. Phys. 1964 , 40 , 213-220.

17. Closs, G. L.; Kaplan, L. R.; Bendall, V. I. J. Am. Chem. Soc. 1967 , 89 , 3376-3377.

18. Closs, G. L.; Boell, W. A.; Heyn, H.; Dev, V. J. Am. Chem. Soc. 1968 , 90 , 173- 178.

217 19. York, E. J.; Dittmar, W.; Stevenson, J. R.; Bergman, R. G. J. Am. Chem. Soc. 1973 , 95 , 5680-5687.

20. Pincock, J. A.; Morchat, R.; Arnold, D. R. J. Amer. Chem. Soc. 1973 , 95 , 7536- 7538.

21. Poutsma, J. C.; Nash, J. J.; Paulino, J. A.; Squires, R. R. J. Am. Chem. Soc. 1997 , 119 , 4686-4697.

22. Yoshimine, M.; Pacansky, J.; Honjou, N. J. Am. Chem. Soc. 1989 , 111 , 2785- 2798.

23. Honjou, N.; Pacansky, J.; Yoshimine, M. J. Am. Chem. Soc. 1985 , 107 , 5332- 5341.

24. Davis, J. H.; Goddard, W. A., III; Bergman, R. G. J. Am. Chem. Soc. 1977 , 99 , 2427-2434.

25. Freeman, P. K. J. Org. Chem. 2009 , 74 , 830-833.

26. Zuev, P. S.; Sheridan, R. S. J. Am. Chem. Soc. 2004 , 126 , 12220-12221.

27. Moss, R. A.; Tian, J.; Sauers, R. R.; Sheridan, R. S.; Bhakta, A.; Zuev, P. S. Org. Lett. 2005 , 7, 4645-4648.

28. Brewbaker, J. L.; Hart, H. J. Amer. Chem. Soc. 1969 , 91 , 711-715.

29. Davies, H. M. L.; Jin, Q. Proc. Natl. Acad. Sci. U. S. A. 2004 , 101 , 5472-5475.

30. Lovely, C. J.; Flores, J. A.; Meng, X.; Dias, H. V. R. Synlett 2009 , 129-132.

31. Doyle, M. P.; Yan, M.; Hu, W.; Gronenberg, L. S. J. Am. Chem. Soc. 2003 , 125 , 4692-4693.

32. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A. J.; T. Vreven; K. N. Kudin; J. C. Burant; J. M. Millam; S. S. Iyengar; J. Tomasi; V. Barone; B. Mennucci; M. Cossi; G. Scalmani; N. Rega, G. A. P., H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al- Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson,

218 W. Chen, M. W. Wong, C. Gonzalez, and J. A. Pople. Gaussian 03, Revision E.01 ed.; Gaussian, Inc.: Wallingford CT, 2004.

33. Wang, J.; Burdzinski, G.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2007 , 129 , 2597-2606.

34. Zhang, Y.; Burdzinski, G.; Kubicki, J.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 16134-16135.

35. Zhang, Y.; Kubicki, J.; Wang, J.; Platz, M. S. J. Phys. Chem. A 2008 , 112 , 11093- 11098.

36. Schrader, T.; Sieg, A.; Koller, F.; Schreier, W.; An, Q.; Zinth, W.; Gilch, P. Chem. Phys. Lett. 2004 , 392 , 358-364.

37. Wang, J.; Burdzinski, G.; Kubicki, J.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 11195-11209.

38. Wang, J.; Kubicki, J.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 2304-2313.

39. Zhang, Y.; Burdzinski, G.; Kubicki, J.; Vyas, S.; Hadad, C. M.; Sliwa, M.; Poizat, O.; Buntinx, G.; Platz, M. S. J. Am. Chem. Soc. 2009 , 131 , 13784-13790.

40. Zhang, Y.; Burdzinski, G.; Kubicki, J.; Platz, M. S. J. Am. Chem. Soc. 2009 , 131 , 9646-9647.

41. Schmitz, C. Ph.D. Thesis, University of Bochum. 1986 .

42. Wang, J.; Burdzinski, G.; Gustafson, T. L.; Platz, M. S. J. Org. Chem. 2006 , 71 , 6221-6228.

43. Zhang, Y.; Burdzinski, G.; Kubicki, J.; Platz, M. S. J. Am. Chem. Soc. 2009 , 131 , 9646-9647.

44. Schwarzer, D.; Troe, J.; Votsmeier, M.; Zerezke, M. J. Chem. Phys. 1996 , 105, 3121-3131.

45. Padwa, A.; Kennedy, G. D. J. Org. Chem. 1984 , 49 , 4344-4352.

46. Kubicki, J.; Zhang, Y.; Wang, J.; Luk, H. L.; Peng, H.-L.; Vyas, S.; Platz, M. S. J. Am. Chem. Soc. 2009 , 131 , 4212-4213.

47. Minato, H. Bull. Chem. Soc. Jpn. 1963 , 36 , 1020-1025.

48. Hess, G. C.; Kohler, B.; Likhotvorik, I.; Peon, J.; Platz, M. S. J. Am. Chem. Soc. 2000 , 122 , 8087-8088.

49. Wang, Y.; Hadad, C. M.; Toscano, J. P. J. Am. Chem. Soc. 2002 , 124 , 1761-1767.

219 50. Wang, Y.; Yuzawa, T.; Hamaguchi, H.-o.; Toscano, J. P. J. Am. Chem. Soc. 1999 , 121 , 2875-2882.

51. Maier, G.; Senger, S. Angew. Chem., Int. Ed. Engl. 1994 , 33 , 558-559.

52. Maier, G.; Lautz, C.; Senger, S. Chem. Eur. J. 2000 , 6, 1467-1473.

53. Albers, R.; Sander, W.; Ottosson, C.-H.; Cremer, D. Chem. Eur. J. 1996 , 2, 967- 973.

54. Scott, A. P.; Platz, M. S.; Radom, L. J. Am. Chem. Soc. 2001 , 123 , 6069-6076.

55. Zhang, Y.; Kubicki, J.; Wang, J.; Platz, M. S. J. Phys. Chem. A 2008 , 112 , 11093- 11098.

56. Scott, A. P.; Radom, L. J. Phys. Chem. 1996 , 100 , 16502-16513.

57. Gonzalez, C.; Schlegel, H. B. J. Chem. Phys. 1989 , 90 , 2154-2161.

58. Gonzalez, C.; Schlegel, H. B. J. Phys. Chem. 1990 , 94 , 5523-5527.

59. Burdzinski, G.; Hackett, J. C.; Wang, J.; Gustafson, T. L.; Hadad, C. M.; Platz, M. S. J. Am. Chem. Soc. 2006, 128 , 13402-13411.

60. Arndt, F. Org. Synth. 1935 , 15 , 3.

220 CHAPTER 8

ULTRAFAST TIME-RESOLVED INFRARED SPECTROSCOPY STUDY ON THE PHOTOCHEMISTRY OF N,N-DETHYLDIAZOACETAMIDE: REARRANGEMENT IN THE EXCITED STATES (RIES)

This chapter is reproduced with permission from J. Am. Chem. Soc. 2008 , 130 , 16134. Copyright 2008 American Chemical Society.

8.1. Introduction

Carbenes often undergo various intramolecular insertion reactions with C−H bonds. 1-3 For example, alkyl- and dialkylcarbenes predominantly insert into the α- and β-

C−H bond to give olefins and cyclopropanes, respectively. 4-7 Formation of four-, five-, and six-membered rings by long-range insertion is possible but rare. 8 For example, photolysis of ethyl diazoacetate in hydrocarbon solvents only gave products from the intermolecular C−H insertion; however, the product of the intramolecular counterpart of this reaction, butyrolactone, was not formed. 9 In contrast, photolytic decomposition of

N,N-diethyldiazoacetamide ( DZA ) in dioxane solution produces β- and γ-lactams in 57% and 43% yields, respectively, exclusively via intramolecular C−H insertion reactions

(Scheme 8.1).10,11

The substantial lactam formation in the photolysis of diazoamides (the absolute overall yield is 94% based on DZA 10 ) has attracted much attention from the synthetic standpoint since the reaction has been shown to be a general method to synthesize

221 families of β-lactam antibiotics, e.g., penicillins,12 and hence many efforts have been made to generate structures containing the β-lactam unit by the diazoamide approach. 13

There has been considerable interest over the years on the mechanism of lactam formation from diazoamides. 10,11,14,15

N2 O h O N N N N H + + 2 O dioxane DZA -lactam -lactam

N2 h O O O H O -lactone Scheme 8.1. Intramolecular reactions in photolytic decomposition of N,N- diethyldiazoacetamide and methyl diazoacetate in solution.

Rando pointed out conformational effects are a dominant influence in these

intramolecular carbene reactions. 10,11 As both diazocarbonyl compounds are planar,16 diazoamides necessarily have an N-alkyl group in proximity to the incipient carbene

center promoting cyclization, but the O-alkyl group of diazoesters is disposed away from

the carbene center, making the alkoxyl group unavailable for cyclization. 11

Rando also found 10 that when DZA was photolytically decomposed in methanol,

the yield of β-lactam was little affected (43%), whereas the formation of γ-lactam is

nearly nullified (5%), with the concomitant formation of the O−H insertion product

(34%). The suppression of γ-lactam formation is analogous to the suppression of

intermolecular C−H insertion when ethyl diazoacetate is photolyzed in methanol. 17 From

these observations, Rando predicted that both lactams were formed via the carbene

222 intermediate, and that the charge separation in their respective transition states were remarkably different. 10,11

Further studies by Tomioka et al . found that the intramolecular C−H insertions in

the photodecomposition of DZA could proceed through the singlet carbene and the

singlet excited state of diazoamide ( 1DZA* ). 14,15 By investigating the effects of sensitizers and quenchers on the product distribution, the authors proposed that 1DZA*

could give rise to the β-lactam directly or through dissociation to nitrogen and singlet carbene, which subsequently undergoes C−H insertion into the C−H bond of the methyl group to give the corresponding γ-lactam ( Scheme 8.2).14 Similar conclusions have been

made in regard to the photochemical Wolff Rearrangement 18 and such processes were

named by Liu as examples of “Rearrangements in Excited States” or RIES. 19

Scheme 8.2. The dual mechanism in β- and γ-lactam formation in the photolysis of N,N- diethyldiazoacetamide in solution.

The dual mechanism in Scheme 8.2 predicts that the β- and γ-lactam will be

formed on different timescales (faster in RIES and slower with the carbene intermediate).

Since the singlet carbene lifetimes are in the sub-nanosecond time scales the formation of

both lactams nicely falls into the measurable time window of ultrafast picosecond time-

223 resolved spectroscopies. In addition, as these two lactams have distinct carbonyl vibrations we were encouraged to investigate the mechanistic details of the photodecomposition of DZA and test the dual mechanistic proposal using ultrafast time-

resolved IR spectroscopy.

8.2. Ultrafast Spectroscopic Results

8.2.1. Ultrafast IR Studies in Chloroform

When one or more photoproducts of interest are persistent, transient spectral

assignments are often facilitated by examining spectral changes induced by photolysis

using conventional FT-IR methods. The spectra recorded before and after photolysis of a

DZA solution in chloroform are shown in Figure 8.1. Upon photolysis, the vibrational

bands due to carbonyl stretching (1602 cm -1) and diazo stretching (2112 cm -1, not shown)

were consumed, and two new bands were formed at 1745 and 1669 cm -1. The 1745 cm -1 band is easily attributed to the β-lactam, due to its characteristically large wavenumber absorption for a carbonyl in a strained four-membered ring, and consistent with prior reports. 10,11 The band at 1699 cm -1 is confidently assigned to the γ-lactam upon comparison with an authentic sample.

224

Figure 8.1. Conventional photolysis of N,N-diethyldiazoacetamide (DZA) with 270 nm laser pulses in chloroform. Stationary FTIR spectra were taken before and after photolysis of a fresh solution in chloroform. The FT-IR spectrum of an authentic sample of γ-lactam is recorded in chloroform. Vibrational frequencies for carbonyl stretches of DZA, the β- and γ-lactams, and singlet carbene in chloroform were obtained from calculations with B3LYP/6-31+G(d) using the PCM model (cf. Table 8.1). The calculated frequencies were scaled by 0.978. The scaling factor was obtained by calibrating the predicted carbonyl frequency of DZA with its experimental value.

Ultrafast time-resolved laser flash photolysis (LFP) of a DZA solution in chloroform with 270 nm laser pulses (300 fs) produced the transient absorption spectra shown in Figure 8.2. In the spectral range expected for carbonyl stretches (1760 – 1650 cm -1), two positive bands due to β-lactam (1745 cm -1) and γ-lactam (1669 cm -1) are detected over a 3 ns time window.

Significantly, these two lactams are formed on two different time scales. Within 2 ps of the laser pulse, β-lactam is detected as a broad band at 1740 – 1680 cm -1, then it narrows and shifts to the blue over 50 ps. These are typical spectral changes observed in

ultrafast IR spectroscopy experiments for newly born species undergoing vibrational

cooling (VC). 20 The β-Lactam is produced within the time resolution of our spectrometer

(< 0.4 ps). The growth time constant (28.9 ± 3.1 ps) obtained by monitoring at 1745 cm -1

225 is thus associated with VC ( Figure 8.3a). After 50 ps, when VC subsides, the β-lactam

band centered at 1745 cm -1 maintains the same intensity over 3 ns ( Figure 8.2b). The instantaneous formation of β-lactam indicates that it is formed from an excited state of the diazoamide precursor ( 1DZA*), confirming the previous proposition by Tomioka. 14,15

Figure 8.2. Transient IR spectra were generated by ultrafast LFP (270 nm) of N,N- diethyldiazoacetamide in chloroform over time windows of (a) 2 – 26 ps and (b) 26 – 1020 ps. The dashed curves are scaled FTIR spectra recorded for an authentic sample of γ-lactam in a chloroform solution.

However, the mechanism of γ-lactam formation is remarkably different. The γ-

lactam band observed at 1669 cm -1 is formed on two time scales: a fast component (~

50%) is already present within 50 ps of the laser pulse when VC subsides, and a slow

component (~ 50%) grows gradually over 50 – 1000 ps ( Figure 8.2b and Figure 8.3b).

The time constant of this slow growth obtained by monitoring at 1669 cm -1 is 190 ± 42 ps

(Figure 8.3b), and is assigned to the γ-lactam formation by cyclization of the singlet

226 carbene ( Scheme 8.2). As will be discussed below, we deduce that both 1DZA* and singlet carbene contribute to the fast component (vide infra ).

Figure 8.3. Kinetic traces obtained by ultrafast IR spectroscopy (270 nm) of N,N- diethyldiazoacetamide in chloroform. (a) The kinetic trace monitored at 1745 cm -1 for β- lactam was fitted to an exponential function. (b) The kinetic trace monitored at 1669 cm -1 for γ-lactam was fitted to an exponential function.

When moving the spectral window to 1680 – 1580 cm -1, a negative band at 1602

cm -1 was observed upon excitation with 270 nm laser pulses due to depletion of DZA concentration ( Figure 8.4). The transient spectra indicate that the ground-state DZA precursor recovered 30% over 100 ps. The shoulder peak at ~ 1640 cm -1 is due to the hot

γ-lactam band undergoing VC.

227

Figure 8.4. Transient IR spectra were generated by ultrafast LFP (270 nm) of N,N- diethyldiazoacetamide (DZA) in chloroform in time windows of 0.7 – 94 ps. The dashed curves are scaled FTIR spectra of γ-lactam and DZA in chloroform.

8.2.2. Ultrafast IR Studies in Methanol-OD

A DZA solution in CH 3OD was bleached with 266 nm light and the resulting FT-

IR spectra are shown in Figure 8.5. The FT-IR spectrum of authentic γ-lactam was also recorded in this solvent for reference purposes. The carbonyl stretching band of the β- lactam was detected at 1750 – 1700 cm -1, and that of the γ-lactam was detected at 1680 –

-1 1630 cm . Both bands were slightly shifted to the red in CH3OD relative to those in chloroform, and were split into doublets. Minato studied the infrared spectra of carbonyl

21 compounds and found that esters and carbonates often have two bands in CH 3OH, one at a wavenumber almost identical with the one in CCl 4, and the other lower by 15 ~ 27 cm -1. These two bands were rationalized as due to mono- and di-hydrogen-bonded carbonyl stretching bands. 21 Even though lactams and amides were not discussed in

Minato’s study, the FT-IR spectra recorded in Figure 8.5 for both lactams and DZA are

21 similar to Minato’s observations, and are attributed to hydrogen bonding with CH 3OD.

228

Figure 8.5. Conventional photolysis (270 nm) of N,N-diethyldiazoacetamide (DZA) in CH 3OD. Stationary FT-IR spectra were taken before and after photolysis of a fresh solution in CH 3OD. The FTIR spectra of AE and authentic γ-lactam were recorded with CH 3OD as the solvent. Vibrational frequencies for carbonyl stretches of DZA, the β- and γ-lactams, amide ether (AE), and singlet carbene in methanol were obtained from calculations with B3LYP/6-31+G(d) using the PCM model (cf. Table 8.1). The calculated frequencies were scaled by 0.978. The scaling factor was obtained by calibrating the predicted carbonyl frequency of DZA with its experimental value.

As is shown in Figure 8.5, the intensity of the β-lactam band remains unchanged,

-1 but that of the γ-lactam band (~ 1670 cm ) is sharply reduced in CH 3OD relative in chloroform. This result is consistent with Rando’s results that γ-lactam yield was depressed by methanol. 10,11 A new band peaking at ~ 1640 cm -1 is observed between γ- lactam and DZA bands. This band is attributed to the amide ether ( AE ) formed by the

11 carbene with CH 3OD (Scheme 8.3). In support of this, AE was synthesized by condensation α-methoxy acetyl chloride with diethylamine, and its FT-IR spectrum is in good agreement with the ~ 1640 cm -1 band formed in photolysis ( Figure

8.5). The vibrational frequencies of AE in methanol were calculated with the polarizable

continuum model (PCM) 22 at the B3LYP/6-31+G(d) level of theory, and were in excellent agreement with the observed 1640 cm -1 band. In addition, the red dashed curve

229 shown in Figure 8.1 clearly reveals the formation of three major products in methanol: β- lactam, γ-lactam and AE . This spectrum was recorded in a chloroform solution of the photoproducts formed upon the photolysis of DZA in CH 3OD (the solvent CH 3OD was

removed from the bleached solution by distillation).

N OMe 2 h MeOD D N N N H H H -N O 2 O O DZA Singlet carbene Amide ether (AE) Scheme 8.3. Formation of singlet carbene from the photolysis of N,N- diethyldiazoacetamide and its reaction with MeOD.

Ultrafast time-resolved transient IR spectra produced with 270 nm laser pulses

(300 fs) in CH 3OD are shown in Figure 8.6 and the generated 2D contour plot is shown in Figure 8.7. The negative band at 1595 cm -1 is due to ground-state DZA bleaching. The

β-lactam band centered at 1733 cm -1 is again formed immediately after the laser pulse, and it has constant intensity between 30 ps and 3 ns. A rising time constant of 5.8 ps is deduced by monitoring at 1733 cm -1 ( Figure 8.8a). This time constant is attributed to

VC, indicating that the VC in CH 3OD is faster than in chloroform.

One of the vibrational bands of the γ-lactam severely overlaps with the amide

ether AE band centered at 1640 cm -1. However, the intensity of γ-lactam probed at 1665 cm -1 remains constant over 10 – 3050 ps ( Figure 8.8b), which is in contrast to the biphasic growth in chloroform (Figure 8.3b). The slow growth over 50 – 1000 ps observed in chloroform is not present in CH 3OD. Thus our data clearly indicate that there are two processes in γ-lactam formation: the fast process which is observed in both chloroform and CH 3OD is from the excited state of DZA via the RIES mechanism, and

230 the slow growth which is only observed in chloroform is from singlet carbene (cf.

Scheme 8.4). The latter mechanism is absent in CH 3OD, indicating that alcohol

efficiently traps the carbene species. 6,7

Figure 8.6. Transient IR spectra were generated by ultrafast LFP (270 nm) of N,N- diethyldiazoacetamide (DZA) in CH 3OD over time windows of (a) 0.7 – 31 ps and (b) 31 – 3050 ps. The dashed curves are FTIR spectra of DZA, AE, and authentic γ-lactam in MeOD.

231

Figure 8.7. A contour plot (2D) of transient IR spectra generated by ultrafast LFP (270 nm) of N,N-diethyldiazoacetamide in CH 3OD in a time window of 0.7 – 3050 ps and spectral window of 1760 – 1580 cm -1.

232

Figure 8.8. Kinetic traces produced upon ultrafast LFP (270 nm) of N,N-diethyl -1 diazoacetamide (DZA) in CH 3OD. (a) The kinetic trace monitored at 1733 cm was fitted to an exponential function. (b) The kinetic trace monitored at 1665 cm -1.

Scheme 8.4. Photochemical reactions for N,N-diethyldiazoacetamide (DZA) in chloroform and CH 3OD from ultrafast time-resolved studies.

It is well accepted that singlet carbonyl carbenes have orthogonal geometries in contrast to their planar precursors. 23 We posit that the orthogonal conformation assists γ-

233 lactam formation because the carbene empty p orbital points towards a γ-CH bond ( vide infra ). Singlet carbene rotation from the nascent planar to the relaxed orthogonal

conformation has been observed for an ester carbene in chloroform,24 consistent with the

current observation that a portion of the γ-lactam is formed from a singlet carbene in

chloroform. However, in methanol solvent molecules surround the nascent singlet

carbene by hydrogen bonding. This provides an intimate solvent shell insulating the

carbene center from the reactive CH bond. Thus, cyclization is suppressed and OH

insertion to solvent is preferred.

Figure 8.9. Kinetic traces monitored at 1706 cm -1 produced from ultrafast LFP (270 nm) of N,N-diethyl diazoacetamide in methanol-OD and was fitted to an exponential function in the time windows of (a) 0.6 – 20 ps and (b) 20 – 3000 ps.

234

Figure 8.10. Kinetic traces monitored at 1643 cm -1 produced from ultrafast IR spectroscopy (270 nm) of N,N-diethyl diazoacetamide in methanol-OD. The kinetic trace at 1643 cm -1 fitted to an exponential function.

It is important to note that another transient species with a band peaking at 1706

-1 cm is observed in CH 3OD ( Figure 8.6 and Figure 8.7). Even though this band overlaps

with one of the β-lactam doublet peaks, it has a decay time constant of 102 ± 12 ps

(Figure 8.9), and remarkably, its decay correlates with the growth of the amide ether

(AE ) band at 1640 cm -1 (109 ± 10 ps, Figure 8.10 ). The carrier of this transient band at

1706 cm -1 is assigned to the singlet carbene based on computational results ( Figure 8.5).

Calculations predict the carbonyl stretching band of singlet carbene at 1709 cm -1 (Table

8.1), in excellent agreement with experimental observations. Another possible candidate

is a cation, formed by protonation of singlet carbene ( Scheme 8.5). However,

calculations predict that the cation absorbs at 1906 cm -1(Table 8.1), exceeding the errors allowed for this calculation and is thus not consistent with the data.

235

Gas Phase Chloroform Methanol Compound Freq (intensity) Freq (intensity) Freq (intensity) carbene 1699 (532) 1707 (783) 1709 (892) γ-lactam 1767 (452) 1712 (760) 1689 (919) β-lactam 1832 (635) 1773 (1033) 1748 (1229) AE 1719 (383) 1687 (605) 1679 (749) cation 1911 (308) 1907 (446) 1906 (505) DZA 1683 (293) 1639 (492) 1620 (606)

Table 8.1. The calculated frequencies (cm -1) for the carbonyl stretching mode in singlet carbene, β-lactam, γ-lactam, AE, cation, and DZA in the gas phase, chloroform and methanol with PCM model at the B3LYP/6-31+G(d) level of theory. Frequencies were not scaled.

Scheme 8.5. Formation of singlet carbene from the photolysis of N,N- diethyldiazoacetamide and its reaction with MeOD. The cation is formed by protonation of singlet carbene.

8.3. Discussions and Calculations

The fact that singlet carbene is detected in CH 3OD but not in chloroform is quite surprising, because singlet carbene is scavengable by methanol but not by chloroform, and thus the opposite result was expected. Ab initio quantum calculations indicate that this observation is fortuitous and caused by the different solvent dependence of carbonyl stretches in lactams and singlet carbene. Calculations with the PCM model predict that in

236 chloroform and methanol a slight blue-shift of the singlet carbene band (by 8 and 10 cm -1,

respectively) and a significant red-shift of the γ-lactam band (by 54 and 78 cm -1, respectively) relative to the gas phase (Table 8.1). The shift of singlet carbene and γ-

lactam bands in opposite directions in these two solvents, and especially the larger shift

and stronger band intensities in methanol relative to chloroform, results in a better

separation between singlet carbene band (1709 cm -1) and γ-lactam band (1689 cm -1) in methanol and hence is consistent with the observation that singlet carbene is less obscured in methanol than in chloroform. 21 In addition, the singlet carbene band is predicted to the blue of γ-lactam by 21 cm -1 in methanol but on the red side in chloroform by 5 cm -1 ( Table 8.1). The switching of the relative positions of these two bands is important in transient IR spectroscopy because the hot ground-state γ-lactam on its red side will obscure the nascent singlet carbene band in chloroform, whereas in methanol the singlet carbene is not affected by VC of γ-lactam because it is on the blue side of the γ- lactam band. Lastly, the faster VC rate in methanol than in chloroform contributes to the clear detection of singlet carbene band in methanol even though it is overlapped with the

β-lactam band. Therefore the predicted trend in the vibrational frequencies nicely correlates with experimental data and supports the conclusion we proposed.

In chloroform at 1669 cm -1, a fast component is present within 50 ps and a slow component grows over 50 – 1000 ps ( τ ~ 190 ps). We have assigned the fast component to γ-lactam formed by RIES. However, it is also possible that the fast component could be singlet carbene that is gradually converted into γ-lactam. The computed IR spectra indicate the overlapping absorptions of singlet carbene (1707 cm -1) and γ-lactam (1717 cm -1, Table 8.1) in chloroform. The carbene absorption is not negligible, as indicated by

237 the results obtained in methanol ( Figure 8.6). Therefore we conclude that the singlet carbene overlaps with γ-lactam and also significantly contributes to the instantaneous rise of the band at 1669 cm -1.

Our data also indicate that β-lactam is entirely formed from the excited state of the diazo amide precursor (RIES), consistent with studies of Tomioka et al .6 It has been

previously shown that the γ-lactam formation is sensitive to methanol quenching; however, it was not clear whether γ-lactam is entirely formed from the carbene pathway.

The observations that some γ-lactam signal is present immediately after excitation in both chloroform and CH 3OD suggest that part of the γ-lactam is also formed from the diazo excited state in chloroform, as well as from relaxed carbene on the ns time scale. The data by Tomioka et al. demonstrated that methanol does not completely suppress γ-lactam formation relative to cyclohexane, indicating that 67% of the γ-lactam is carbene-derived, and 33% is formed by RIES. 14,15 If we assume that the extinction coefficients of γ-lactam

and the amide ether (AE) are similar, as supported by calculations in Table 8.1, the yield

of γ-lactam formed from the RIES pathway is about 30% (see transient absorption

spectrum at 3050 ps in Figure 8.6), which is consistent with the results by Tomioka et

al .14,15 In addition, the greater faster component observed in chloroform is consistent with

the proposition that singlet carbene is formed in chloroform and overlaps with γ-lactam.

238 6.2

TS2 4.4

TS1 0

Singlet Carbene

-75.2

β-lactam

-147.5

γ-lactam

Figure 8.11. The transition states for the formation of γ-lactam (TS1) and β-lactam (TS2) from singlet carbene calculated at the B3LYP/6-31+G(d) level of theory. Free energies (∆G) are in kcal/mol at 298 K in the gas phase. ZPE was corrected.

The transition states for β-lactam and γ-lactam formation from the singlet carbene were located with B3LYP/6-31+G(d) calculations ( Figure 8.11 , Table 8.2 and Table

8.3). The activation energy for γ-lactam formation ( TS1 , 4.4 kcal/mol) is predicted to be

1.8 kcal/mol less than that of β-lactam ( TS2 , 6.2 kcal/mol), consistent with the selective formation of γ-lactam from singlet carbene. β-Lactam is considerably less stable than that of the γ-lactam (higher in energy by 72.3 kcal/mol), reflecting the excessive strain associated with β-lactam. Therefore selective formation of γ-lactam is thermodynamically more favorable. Transition state theory predicts that the lifetime of the relaxed singlet carbene is 271 ps at 298 K. This is in good agreement with the 190 ±

42 ps growth of γ-lactam in chloroform.

239 The singlet carbene is predicted to have an orthogonal geometry ( Figure 8.11 and

Table 8.4), in contrast to the planar precursors and the triplet state, consistent with previous results.23 Therefore, we posit that the intramolecular CH insertion is facilitated by overlapping the carbene empty p orbital with H atoms of an N-alkyl group. Even

though the α-CH of the N-alkyl group is much closer than the β-CH to the carbene empty p orbital, the β-lactam is not formed because its transition state ( TS2 ) is higher than that of γ-lactam ( TS1 ) and is less exothermic than γ-lactam formation.

The details of the RIES mechanism leading to lactams are intriguing and it is

rather hard to describe a scheme showing direct C-H insertion in the excited state leading

to lactams. It is possible that the diazo excited state abstracts a hydrogen atom from a CH

bond of the N-ethyl group to form a singlet biradical which cyclizes to form the two

lactams. The putative biradicals were not observed and such species usually can not be

located as minima on potential energy surfaces. A full excited state computational study

is needed in the future to understand this mechanism. Alternatively, we posit that the

RIES mechanism could also be a concerted process, in which the hydrogen atom

migration is concerted with the nitrogen molecule dissociation, similar to the concerted

Wolff rearrangement. 18,25 In this regard, we note that hydrogen atom migration is assisted by overlapping the π* antibonding orbital of the adjacent N 2 in β-lactam formation, while this orbital overlapping is not favorable in the formation of γ-lactam ( Table 8.5).

240

E = -365.12479850 Hartree C 0.46256 0.98237 -0.01857 C 1.78983 1.02794 0.56257 H 2.51525 1.4081 -0.17104 O -0.02886 2.12431 -0.11908 N -0.18383 -0.18356 -0.31392 C 0.46825 -1.4429 0.03378 H 0.20329 -1.75316 1.05643 H 0.10856 -2.22349 -0.64831 C 1.98148 -1.29482 -0.07683 H 2.51363 -2.15585 0.34245 H 2.32116 -1.10582 -1.09705 H 2.30731 -0.44333 0.62646 C -1.61427 -0.1622 -0.62222 H -1.79739 -0.90736 -1.40733 H -1.82763 0.82502 -1.03958 C -2.51594 -0.42529 0.58956 H -2.34857 0.32882 1.36606 H -3.56904 -0.37664 0.28772 H -2.34042 -1.41653 1.02453 Frequency IR intensity Frequency IR intensity Frequency IR intensity (cm -1) (km/mol) (cm -1) (km/mol) (cm -1) (km/mol) -298 82.3 948 6.1 1502 35.0 60 2.6 964 24.3 1517 13.8 111 2.2 1021 31.1 1523 2.8 171 0.3 1060 22.5 1542 27.3 183 1.8 1099 9.1 1558 6.5 219 1.3 1110 4.8 1632 311.2 324 13.7 1159 15.1 2333 148.8 358 5.5 1207 7.6 3010 44.6 390 15.5 1261 13.3 3047 30.9 479 11.3 1300 3.9 3048 39.2 513 0.5 1342 118.7 3064 26.7 596 19.9 1368 6.1 3086 8.3 632 23.5 1401 22.0 3092 20.7 702 3.8 1407 10.3 3105 22.1 794 3.9 1432 11.9 3123 23.5 830 0.9 1468 1.6 3136 13.6 882 25.2 1492 94.1 3163 9.4

Table 8.2. Optimized structure of the transition state TS1 from singlet carbene to γ- lactam at the B3LYP/6-31+G(d) level of theory.

241

E = -365.11974372 Hartree C 1.350821000 -0.257373000 -0.058775000 C 2.031504000 0.980121000 0.319077000 H 2.350820000 0.956168000 1.374446000 O 1.997763000 -1.271881000 -0.332535000 N -0.004020000 -0.063473000 -0.137501000 C -0.323678000 1.233151000 0.419142000 H -0.589367000 1.191388000 1.483449000 H 0.709776000 1.790311000 0.364752000 C -1.293673000 2.085304000 -0.387299000 H -2.292871000 1.632454000 -0.392499000 H -1.384211000 3.081929000 0.057313000 H -0.954569000 2.190234000 -1.422906000 C -0.961281000 -1.073745000 -0.563091000 H -0.366682000 -1.851589000 -1.051950000 H -1.626392000 -0.643812000 -1.323300000 C -1.776927000 -1.672981000 0.588134000 H -2.381291000 -0.913379000 1.099280000 H -2.461541000 -2.438589000 0.204212000 H -1.118243000 -2.142614000 1.326859000 Frequency IR intensity Frequency IR intensity Frequency IR intensity (cm -1) (km/mol) (cm -1) (km/mol) (cm -1) (km/mol) -303 40.0 955 20.8 1503 16.1 51 0.4 978 3.5 1512 9.0 98 3.6 996 50.1 1516 6.9 117 2.7 1055 1.5 1517 2.3 190 1.8 1096 8.2 1526 5.2 210 0.1 1107 1.5 1710 378.4 258 4.4 1135 5.2 2154 193.0 300 4.4 1194 1.1 3045 16.6 327 6.0 1261 8.2 3046 21.6 434 7.0 1342 41.7 3053 2.1 498 8.4 1361 42.0 3053 47.6 592 29.9 1395 34.8 3056 31.1 614 17.0 1403 8.2 3101 15.1 722 1.3 1424 2.3 3116 18.8 763 5.0 1431 18.2 3118 21.1 791 3.2 1434 10.6 3127 21.4 882 25.2 1492 94.1 3163 9.4

Table 8.3. Optimized structure of the transition state TS2 from singlet carbene to β- lactam at the B3LYP/6-31+G(d) level of theory.

242

E = -365.1309144 Hartree C 1.149804000 -0.371627000 -0.007924000 C 1.941001000 -1.360028000 0.609092000 H 2.209739000 -2.148219000 -0.108165000 O 2.161311000 0.240553000 -0.477386000 N -0.128469000 0.009111000 -0.022072000 C -1.141637000 -0.849701000 0.598330000 H -0.623928000 -1.496837000 1.315945000 H -1.818895000 -0.211796000 1.178178000 C -1.930060000 -1.690600000 -0.410327000 H -2.675677000 -2.301491000 0.112224000 H -2.459892000 -1.060720000 -1.133789000 H -1.265108000 -2.361252000 -0.965406000 C -0.498994000 1.310146000 -0.596046000 H -1.441508000 1.181486000 -1.140126000 H 0.270448000 1.570482000 -1.328186000 C -0.623714000 2.416908000 0.455835000 H 0.330905000 2.569620000 0.969464000 H -0.909197000 3.359078000 -0.027022000 H -1.386482000 2.180861000 1.206714000 Frequency IR intensity Frequency IR intensity Frequency IR intensity (cm -1) (km/mol) (cm -1) (km/mol) (cm -1) (km/mol) 52 0.6 934 70.4 1511 20.4 67 2.3 955 63.1 1515 1.1 127 2.1 972 3.5 1517 3.9 137 3.9 1067 29.7 1525 3.6 181 1.0 1094 12.0 1527 12.9 205 0.3 1106 7.7 1698 520.3 253 5.2 1131 13.8 3050 16.3 309 23.2 1225 15.8 3051 12.3 321 9.5 1255 5.5 3057 33.2 410 13.6 1337 2.5 3069 27.3 436 26.7 1378 29.1 3076 15.7 540 22.0 1395 8.0 3094 3.8 582 7.8 1418 23.6 3109 14.5 702 1.0 1433 6.3 3118 31.3 793 4.9 1437 25.0 3123 17.9 799 8.3 1462 65.4 3127 29.3 868 18.2 1505 3.4 3138 27.3

Table 8.4. Optimized structure of the singlet carbene at the B3LYP/6-31+G(d) level of theory.

243

E = -474.8351868 Hartree C 0.472977000 -0.385961000 -0.159551000 C 1.534449000 0.544322000 0.268959000 H 1.432241000 1.563319000 0.603584000 O 0.753796000 -1.492328000 -0.615536000 N 2.762028000 0.104530000 0.180089000 N 3.812033000 -0.301336000 0.095869000 N -0.822787000 0.050097000 -0.031189000 C -1.217478000 1.324301000 0.570272000 H -2.183012000 1.166399000 1.058969000 H -0.522929000 1.578370000 1.373913000 C -1.329416000 2.477764000 -0.432678000 H -1.663521000 3.389727000 0.070914000 H -0.368448000 2.682381000 -0.910875000 H -2.050555000 2.243703000 -1.219826000 C -1.891292000 -0.838394000 -0.505620000 H -1.494367000 -1.416880000 -1.339189000 H -2.698639000 -0.207672000 -0.889728000 C -2.416818000 -1.782026000 0.578697000 H -2.809224000 -1.230509000 1.438240000 H -3.226020000 -2.401007000 0.180193000 H -1.619341000 -2.442276000 0.924224000 Frequency IR intensity Frequency IR intensity Frequency IR intensity (cm -1) (km/mol) (cm -1) (km/mol) (cm -1) (km/mol) 54 0.6 790 3.5 1480 41.3 71 0.2 844 7.6 1491 2.3 80 4.0 924 3.2 1493 7.7 92 4.2 957 9.8 1500 7.7 134 0.6 1024 1.0 1504 12.2 154 0.6 1086 13.7 1520 27.3 216 0.1 1098 6.6 1674 301.2 225 1.0 1114 18.7 2218 641.4 295 1.7 1147 113.1 3026 24.7 329 2.8 1204 44.8 3028 17.3 385 6.3 1245 22.8 3037 38.1 427 5.2 1277 98.0 3046 26.9 452 23.1 1338 13.7 3074 4.9 456 19.5 1380 14.7 3082 24.3 533 2.2 1388 125.6 3094 37.3 554 0.4 1401 11.6 3104 36.5 624 4.6 1410 7.6 3107 23.4 738 22.4 1413 21.9 3120 16.1 782 6.9 1454 385.2 3246 2.3

Table 8.5. Optimized structure of the ground state N,N-diethyldiazoacetamide at the B3LYP/6-31+G(d) level of theory.

244 8.4. Conclusions

In summary, our experiments indicate that in chloroform, both β- and γ -lactams are formed from the diazo amide precursor via RIES, and that γ-lactam is also formed by isomerization of relaxed singlet carbene. In methanol both carbene decay and the rise of amide ether product are observed directly. The predictions of DFT calculations are consistent with the experimental observations.

8.5. Computational and Experimental Details

8.5.1. Calculations

Calcualtions were performed using Becke’s three-parameter hybrid exchange functional with the Lee-Yang-Parr correlation functional (B3LYP) 26-29 as implemented in

Gaussian 03.30 Geometries were optimized at the B3LYP/6-31G+(d) level of theory.

Vibrational frequency analyses at the B3LYP/6-31G+(d) level were utilized to verify that

stationary points obtained corresponded to energy minima. In order to approximate the

experimental condensed phase and to evaluate the effect of different solvents, the

polarizable continuum model (PCM) 22 implemented in Gaussian 03 was utilized.

Transition states were located with the STQN method31 implemented in Gaussian 03. The zero point energy (ZPE) was corrected with a factor of 0.9806. All calculations were performed at The Ohio Supercomputer Center.

8.5.2. Ultrafast Spectroscopy

Ultrafast IR pump-probe absorption measurements were performed using the home-built spectrometer at the Ohio State University. Solution concentrations were adjusted to absorption of unity in a 1 mm cell. Sample solutions were excited in a

245 stainless steel flow cell equipped with 2 mm thick BaF2 windows. After passing the sample reference and probe beam were spectrally dispersed with a polychromator and independently imaged on a liquid-nitrogen cooled HgCdTe detector (2 x 32 pixels). The pump pulse energy was about 4 uJ at the sample position. The entire set of pump-probe delay positions (cycle) is repeated at least three times, to observe data reproducibility from cycle to cycle. To avoid rotational diffusion effects, the angle between polarizations of the pump beam and the probe beam was set to the magic angle (54.7°). Kinetic traces are analyzed by fitting to a sum of exponential terms. All experiments were performed at room temperature.

8.5.3. Synthesis and Materials

p-Nitrophenyl diazoacetate was prepared and purified by the published procedure. 17,32 N,N-Diethyldiazoacetamide was prepared by reacting p-nitrophenyl diazoacetate with diethylamine in ether as previously reported, 10,11,14,15 and was purified

on alumina (basic) column. l-Ethyl-2-pyrollidinone ( γ-lactam) was purchased from

Sigma-Aldrich, Inc. N,N-Diethyl-α-methoxyacetamide was prepared by reacting α-

methoxyacetyl chloride with excess diethylamine in dry ether. All other materials and

solvents were purchased from Sigma-Aldrich, Inc.

8.6. References for Chapter 8

1. Kirmse, W. Carbene Chemistry, 2nd Ed. ; Academic Press: New York, 1971.

2. Moss, R. A.; Platz, M. S.; Jones, M., Jr. Reactive Intermediate Chemistry ; Wiley: New York, 2004.

3. Moss, R. A.; Turro, N. J. In Kinetics and Spectroscopy of Carbenes and Biradicals ; Platz, M. S., Ed.; Plenum: New York, 1990; p 213f.

4. Frey, H. M. Pure and Applied Chemistry 1964 , 9, 527-537.

246 5. Friedman, L.; Shechter, H. J. Am. Chem. Soc. 1959 , 81 , 5512-5513.

6. Mansoor, A. M.; Stevens, I. D. R. Tetrahedron Lett. 1966 , 1733-1737.

7. Closs, G. L. J. Am. Chem. Soc. 1962 , 84 , 809-813.

8. Friedman, L.; Shechter, H. J. Am. Chem. Soc. 1961 , 83 , 3159-3160.

9. Kirmse, W.; Buecking, H. W.; Buecking, H. W. Tetrahedron Lett. 1967 , 1833- 1835.

10. Rando, R. R. J. Am. Chem. Soc. 1970 , 92 , 6706-6707.

11. Rando, R. R. J. Am. Chem. Soc. 1972 , 94 , 1629-1631.

12. Corey, E. J.; Felix, A. M. J. Am. Chem. Soc. 1965 , 87 , 2518-2519.

13. Candeias, N. R.; Gois, P. M. P.; Veiros, L. F.; Afonso, C. A. M. J. Org. Chem. 2008 , 73 , 5926-5932.

14. Tomioka, H.; Kitagawa, H.; Izawa, Y. J. Org. Chem. 1979 , 44 , 3072-3075.

15. Tomioka, H.; Kondo, M.; Izawa, Y. J. Org. Chem. 1981 , 46 , 1090-1094.

16. Kaplan, F.; Meloy, G. K. J. Am. Chem. Soc. 1966 , 88 , 950-956.

17. Chaimovich, H.; Vaughan, R. J.; Westheimer, F. H. J. Am. Chem. Soc. 1968 , 90 , 4088-4093.

18. Burdzinski, G. T.; Wang, J.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 3746-3747.

19. Bonneau, R.; Liu, M. T. H.; Kim, K. C.; Goodman, J. L. J. Am. Chem. Soc. 1996 , 118 , 3829-3837.

20. Hamm, P.; Ohline, S. M.; Zinth, W. J. Chem. Phys. 1997 , 106 , 519-529.

21. Minato, H. Bull. Chem. Soc. Jpn. 1963 , 36 , 1020-1025.

22. Mennucci, B.; Tomasi, J. J. Chem. Phys. 1997 , 106 , 5151-5158.

23. Scott, A. P.; Platz, M. S.; Radom, L. J. Am. Chem. Soc. 2001 , 123 , 6069-6076.

24. Zhang, Y.; Kubicki, J.; Wang, J.; Platz, M. S. J. Phys. Chem. A 2008 , 112 , 11093- 11098.

25. Wang, J.; Burdzinski, G.; Kubicki, J.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 11195-11209.

247 26. Becke, A. D. J. Chem. Phys. 1993 , 98 , 5648-5652.

27. Becke, A. D. J. Chem. Phys. 1992 , 96 , 2155-2160.

28. Becke, A. D. J. Chem. Phys. 1992 , 97 , 9173-9177.

29. Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988 , 37 , 785-789.

30. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A. J.; T. Vreven; K. N. Kudin; J. C. Burant; J. M. Millam; S. S. Iyengar; J. Tomasi; V. Barone; B. Mennucci; M. Cossi; G. Scalmani; N. Rega, G. A. P., H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al- Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez, and J. A. Pople. Gaussian 03, Revision E.01 ed.; Gaussian, Inc.: Wallingford CT, 2004.

31. Peng, C.; Ayala, P.; Schlegel, H. B.; Frisch, M. J. J. Comput. Chem. 1996 , 17 , 49- 56.

32. Shafer, J.; Barnowsky, P.; Laursen, R.; Finn, F.; Westheimer, F. H. J. Biol. Chem. 1966 , 241 , 421-427.

248 BIBLIOGRAPHY

1. IUPAC Gold Book Website. http://goldbook.iupac.org/C00806.html (accessed February 4, 2010).

2. Kirmse, W. Carbene Chemistry, 2nd Ed. ; Academic Press: New York, 1971.

3. Dumas, J. B. Ann. Chim. Phys. 1835 , 58 , 28.

4. Regnault, H. V. Ann. Chim. Phys. 1839 , 71 , 427.

5. Butlerov, A. M. Ann. 1861 , 120 , 356.

6. Geuther, A. Ann. Chem. Pharm. 1862 , 123 , 121-122.

7. Platz, M. S. Observing Invisible Carbenes By Trapping Them with Pyridine. In Carbene Chemistry: From Fleeting Intermediates to Powerful Reagents ; Bertrand, G., Ed.; Marcel Dekker: New York, 2002; pp 27-56.

8. Buchner, E.; Feldmann, L. Ber. Dtsch. chem. Ges. 1903 , 36 , 3509-3517.

9. Staudinger, H.; Kupfer, O. Ber. Dtsch. Chem. Ges. 1912 , 45 , 501-509.

10. Doering, W. v. E.; Knox, L. H. J. Am. Chem. Soc. 1956 , 78 , 4947-4950.

11. Hine, J. J. Am. Chem. Soc. 1950 , 72 , 2438-2445.

12. Doering, W. v. E.; Hoffmann, A. K. J. Am. Chem. Soc. 1954 , 76 , 6162-6165.

13. Doering, W. v. E.; Buttery, R. G.; Laughlin, R. G.; Chaudhuri, N. J. Am. Chem. Soc. 1956 , 78 , 3224.

14. Herzberg, G.; Shoosmith, J. Nature 1959 , 183 , 1801-1802.

15. Nobel Foundation Website (1971). http://nobelprize.org/nobel_prizes/chemistry/laureates/1971/index.html (accessed February 4, 2010).

249 16. Woodworth, R. C.; Skell, P. S. J. Am. Chem. Soc. 1959 , 81 , 3383-3386.

17. Whittle, E.; Dows, D. A.; Pimentel, G. C. J. Chem. Phys. 1954 , 22 , 1943.

18. Becker, E. D.; Pimentel, G. C. J. Chem. Phys. 1956 , 25 , 224-228.

19. Norman, I.; Porter, G. Nature 1954 , 174 , 508-509.

20. Murray, R. W.; Trozzolo, A. M.; Wasserman, E.; Yager, W. A. J. Am. Chem. Soc. 1962 , 84 , 3213-3214.

21. Gibbons, W. A.; Trozzolo, A. M. J. Am. Chem. Soc. 1966 , 88 , 172-173.

22. Trozzolo, A. M.; Murray, R. W.; Wasserman, E. J. Am. Chem. Soc. 1962 , 84 , 4990-4991.

23. Trozzolo, A. M.; Gibbons, W. A. J Am Chem Soc 1967 , 89 , 239-243.

24. Ware, W. R.; Sullivan, P. J. J. Chem. Phys. 1968 , 49 , 1445-1446.

25. Closs, G. L.; Hutchisin, C. A., Jr.; Kohler, B. E. J. Chem. Phys. 1966 , 44 , 413-414.

26. Brandon, R. W.; Closs, G. L.; Hutchison, C. A., Jr. J. Chem. Phys. 1962 , 37 , 1878-1879.

27. Paulsen, S. R. Angew. Chem. 1960 , 72 , 781-782.

28. Schmitz, E.; Ohme, R. Chem. Ber. 1961 , 94 , 2166-2173.

29. Schmitz, E.; Ohme, R. Tetrahedron Lett. 1961 , 612-614.

30. Goddard, W. A., III. Science 1985 , 227 , 917-923.

31. Schaefer, H. F., III. Science 1986 , 231 , 1100-1107.

32. Foster, J. M.; Boys, S. F. Rev. Mod. Phys. 1960 , 32 , 305-307.

33. Harding, L. B.; Goddard, W. A., III. Chem. Phys. Lett. 1978 , 55 , 217-220.

34. Zittel, P. F.; Ellison, G. B.; Oneil, S. V.; Herbst, E.; Lineberger, W. C.; Reinhardt, W. P. J. Am. Chem. Soc. 1976 , 98 , 3731-3732.

35. Nobel Foundation Website (1967). http://nobelprize.org/nobel_prizes/chemistry/laureates/1967/ (accessed February 4, 2010).

36. Small, R. D., Jr.; Scaiano, J. C. J. Am. Chem. Soc. 1978 , 100 , 296-298.

250 37. Closs, G. L.; Rabinow, B. E. J. Am. Chem. Soc. 1976 , 98 , 8190-8198.

38. Brauer, B. E.; Grasse, P. B.; Kaufmann, K. J.; Schuster, G. B. J. Am. Chem. Soc. 1982 , 104 , 6814-6816.

39. Zupancic, J. J.; Schuster, G. B. J. Am. Chem. Soc. 1980 , 102 , 5958-5960.

40. Grasse, P. B.; Brauer, B. E.; Zupancic, J. J.; Kaufmann, K. J.; Schuster, G. B. J. Am. Chem. Soc. 1983 , 105 , 6833-6845.

41. Griller, D.; Montgomery, C. R.; Scaiano, J. C.; Platz, M. S.; Hadel, L. J. Am. Chem. Soc. 1982 , 104 , 6813-6814.

42. Hadel, L. M.; Platz, M. S.; Scaiano, J. C. J. Am. Chem. Soc. 1984 , 106 , 283-287.

43. Hadel, L. M.; Maloney, V. M.; Platz, M. S.; McGimpsey, W. G.; Scaiano, J. C. J. Phys. Chem. 1986 , 90 , 2488-2491.

44. Barcus, R. L.; Hadel, L. M.; Johnston, L. J.; Platz, M. S.; Savino, T. G.; Scaiano, J. C. J. Am. Chem. Soc. 1986 , 108 , 3928-3937.

45. Horn, K. A.; Chateauneuf, J. E. Tetrahedron 1985 , 41 , 1465-1470.

46. Moss, R. A.; Turro, N. J. In Kinetics and Spectroscopy of Carbenes and Biradicals ; Platz, M. S., Ed.; Plenum: New York, 1990; pp 213f.

47. Gould, I. R.; Turro, N. J.; Butcher, J., Jr.; Doubleday, C., Jr.; Hacker, N. P.; Lehr, G. F.; Moss, R. A.; Cox, D. P.; Guo, W.; et al. Tetrahedron 1985 , 41 , 1587-1600.

48. Turro, N. J.; Butcher, J. A.; Moss, R. A.; Guo, W.; Munjal, R. C.; Fedorynski, M. J. Am. Chem. Soc. 1980 , 102 , 7576-7578.

49. Jackson, J. E.; Soundararajan, N.; Platz, M. S.; Liu, M. T. H. J. Am. Chem. Soc. 1988 , 110 , 5595-5596.

50. Modarelli, D. A.; Platz, M. S. J. Am. Chem. Soc. 1991 , 113 , 8985-8986.

51. Modarelli, D. A.; Morgan, S.; Platz, M. S. J. Am. Chem. Soc. 1992 , 114 , 7034- 7041.

52. Modarelli, D. A.; Platz, M. S.; Sheridan, R. S.; Ammann, J. R. J. Am. Chem. Soc. 1993 , 115 , 10440-10441.

53. Ford, F.; Yuzawa, T.; Platz, M. S.; Matzinger, S.; Fuelscher, M. J. Am. Chem. Soc. 1998 , 120 , 4430-4438.

251 54. Pezacki, J. P.; Couture, P.; Dunn, J. A.; Warkentin, J.; Wood, P. D.; Lusztyk, J.; Ford, F.; Platz, M. S. J. Org. Chem. 1999 , 64 , 4456-4464.

55. Liu, M. T. H.; Bonneau, R. J. Am. Chem. Soc. 1989 , 111 , 6873-6874.

56. Jackson, J. E.; Platz, M. S. Laser flash photolysis studies of ylide-forming reactions of carbenes. In Advances in Carbene Chemistry ; Brinker, U. H., Ed.; JAI Press: Greenwich CT, 1994; pp 89-160.

57. Zewail, A. H. Angew. Chem. Int. Ed. 2000 , 39 , 2586-2631.

58. Zewail, A. H. J. Phys. Chem. A 2000 , 104 , 5660-5694.

59. Nobel Foundation Website (1999). http://nobelprize.org/nobel_prizes/chemistry/laureates/1999/index.html (accessed February 4, 2010).

60. Peon, J.; Polshakov, D.; Kohler, B. J. Am. Chem. Soc. 2002 , 124 , 6428-6438.

61. Wang, J.; Kubicki, J.; Hilinski, E. F.; Mecklenburg, S. L.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2007 , 129 , 13683-13690.

62. Wang, J.; Burdzinski, G.; Gustafson, T. L.; Platz, M. S. J. Org. Chem. 2006 , 71 , 6221-6228.

63. Wang, J.; Burdzinski, G.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2007 , 129 , 2597-2606.

64. Mueller-Remmers, P. L.; Jug, K. J. Am. Chem. Soc. 1985 , 107 , 7275-7284.

65. Bigot, B.; Ponec, R.; Sevin, A.; Devaquet, A. J. Am. Chem. Soc. 1978 , 100 , 6575- 6580.

66. Yamamoto, N.; Bernardi, F.; Bottoni, A.; Olivucci, M.; Robb, M. A.; Wilsey, S. J. Am. Chem. Soc. 1994 , 116 , 2064-2074.

67. Bernardi, F.; Olivucci, M.; Robb, M. A.; Vreven, T.; Soto, J. J. Org. Chem. 2000 , 65 , 7847-7857.

68. Arenas, J. F.; Lopez-Tocon, I.; Otero, J. C.; Soto, J. J. Am. Chem. Soc. 2002 , 124 , 1728-1735.

69. Wang, J.; Burdzinski, G.; Kubicki, J.; Platz, M. S.; Moss, R. A.; Fu, X.; Piotrowiak, P.; Myahkostupov, M. J. Am. Chem. Soc. 2006 , 128 , 16446-16447.

70. Zhang, Y.; Burdzinski, G.; Kubicki, J.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 16134-16135. 252 71. Zhang, Y.; Burdzinski, G.; Kubicki, J.; Vyas, S.; Hadad, C. M.; Sliwa, M.; Poizat, O.; Buntinx, G.; Platz, M. S. J. Am. Chem. Soc. 2009 , 131 , 13784-13790.

72. Zhang, Y.; Wang, L.; Moss, R. A.; Platz, M. S. J. Am. Chem. Soc. 2009 , 131 , 16652-16653.

73. Zhang, Y.; Kubicki, J.; Platz, M. S. J. Am. Chem. Soc. 2009 , 131 , 13602-13603.

74. Zhang, Y.; Burdzinski, G.; Kubicki, J.; Platz, M. S. J. Am. Chem. Soc. 2009 , 131 , 9646-9647.

75. Moss, R. A.; Platz, M. S.; Jones, M., Jr. Reactive Intermediate Chemistry ; Wiley: New York, 2004.

76. Liu, M. T. H. Chemistry of Diazirines ; CRC Press, 1987.

77. Moss, R. A. Acc. Chem. Res. 2006 , 39 , 267-272.

78. Jennings, B. M.; Liu, M. T. H. J. Am. Chem. Soc. 1976 , 98 , 6416-6417.

79. Wang, J.; Kubicki, J.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 2304-2313.

80. Wang, J.; Burdzinski, G.; Kubicki, J.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 11195-11209.

81. Wang, J.; Kubicki, J.; Hilinski, E. F.; Mecklenburg, S. L.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2007 , 129 , 13683-13690.

82. Nigam, M.; Platz, M. S.; Showalter, B. M.; Toscano, J. P.; Johnson, R.; Abbot, S. C.; Kirchhoff, M. M. J. Am. Chem. Soc. 1998 , 120 , 8055-8059.

83. Bonneau, R.; Liu, M. T. H. J. Am. Chem. Soc. 1996 , 118 , 7229-7230.

84. Bonneau, R.; Liu, M. T. H.; Kim, K. C.; Goodman, J. L. J. Am. Chem. Soc. 1996 , 118 , 3829-3837.

85. Liu, M. T. H.; Jenning, B. M. Can. J. Chem. 1977 , 55 , 6.

86. Chandrasekhar, S. Res. Chem. Intermed. 1992 , 17 , 173-209.

87. Skell, P. S.; Garner, A. Y. J. Am. Chem. Soc. 1956 , 78 , 3409-3411.

88. Hoffmann, R. J. Am. Chem. Soc. 1968 , 90 , 1475-1485.

89. Zurawski, B.; Kutzelnigg, W. J. Am. Chem. Soc. 1978 , 100 , 2654-2659.

253 90. Rondan, N. G.; Houk, K. N.; Moss, R. A. J. Am. Chem. Soc. 1980 , 102 , 1770- 1776.

91. Houk, K. N.; Rondan, N. G. J. Am. Chem. Soc. 1984 , 106 , 4293-4294.

92. Blake, J. F.; Wierschke, S. G.; Jorgensen, W. L. J. Am. Chem. Soc. 1989 , 111 , 1919-1920.

93. Keating, A. E.; Garcia-Garibay, M. A.; Houk, K. N. J. Am. Chem. Soc. 1997 , 119 , 10805-10809.

94. Bernardi, F.; Bottoni, A.; Canepa, C.; Olivucci, M.; Robb, M. A.; Tonachini, G. J. Org. Chem. 1997 , 62 , 2018-2025.

95. Keating, A. E.; Merrigan, S. R.; Singleton, D. A.; Houk, K. N. J. Am. Chem. Soc. 1999 , 121 , 3933-3938.

96. Becke, A. D. J. Chem. Phys. 1993 , 98 , 5648-5652.

97. Becke, A. D. J. Chem. Phys. 1992 , 96 , 2155-2160.

98. Becke, A. D. J. Chem. Phys. 1992 , 97 , 9173-9177.

99. Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988 , 37 , 785-789.

100. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A. J.; T. Vreven; K. N. Kudin; J. C. Burant; J. M. Millam; S. S. Iyengar; J. Tomasi; V. Barone; B. Mennucci; M. Cossi; G. Scalmani; N. Rega, G. A. P., H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al- Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez, and J. A. Pople. Gaussian 03, Revision E.01 ed.; Gaussian, Inc.: Wallingford CT, 2004.

101. Hattig, C. J. Chem. Phys. 2003 , 118 , 7751-7761.

102. Haettig, C.; Koehn, A. J. Chem. Phys. 2002 , 117 , 6939-6951.

103. Kohn, A.; Hattig, C. J. Chem. Phys. 2003 , 119 , 5021-5036.

254 104. Ahlrichs, R.; Baer, M.; Haeser, M.; Horn, H.; Koelmel, C. Chem. Phys. Lett. 1989 , 162 , 165-169.

105. Scott, A. P.; Radom, L. J. Phys. Chem. 1996 , 100 , 16502-16513.

106. Molecular graphics images were produced using the UCSF Chimera package from the Resource for Biocomputing, Visualization, and Informatics at the University of California, San Francisco (supported by NIH P41 RR-01081)

107. Pettersen, E. F.; Goddard, T. D.; Huang, C. C.; Couch, G. S.; Greenblatt, D. M.; Meng, E. C.; Ferrin, T. E. J. Comput. Chem. 2004 , 25 , 1605-1612.

108. Gonzalez, C.; Schlegel, H. B. J. Chem. Phys. 1989 , 90 , 2154-2161.

109. Gonzalez, C.; Schlegel, H. B. J. Phys. Chem. 1990 , 94 , 5523-5527.

110. Reed, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys. 1985 , 83 , 735-746.

111. Wang, J.; Kubicki, J.; Burdzinski, G.; Hackett, J. C.; Gustafson, T. L.; Hadad, C. M.; Platz, M. S. J. Org. Chem. 2007 , 72 , 7581-7586.

112. Burdzinski, G.; Hackett, J. C.; Wang, J.; Gustafson, T. L.; Hadad, C. M.; Platz, M. S. J. Am. Chem. Soc. 2006, 128 , 13402-13411.

113. Laube, T.; Olah, G. A.; Bau, R. J. Am. Chem. Soc. 1997 , 119 , 3087-3092.

114. Turro, N. J. Modern Molecular Photochemistry ; University Press: Menlo Park, CA, 1978.

115. Frey, H. M.; Stevens, I. D. R. J. Chem. Soc. 1962 , 3865-3867.

116. Zhang, Y.; Platz, M. S. Org. Lett. 2009 , submitted .

117. Liu, M. T. H.; Ramakrishnan, K. J. Org. Chem. 1977 , 42 , 3450-3452.

118. Graham, W. H. J. Am. Chem. Soc. 1962 , 84 , 1063-1064.

119. Graham, W. H. J. Org. Chem. 1965 , 30 , 2108.

120. Graham, W. H. J. Am. Chem. Soc. 1966 , 88 , 4677-4681.

121. Graham, W. H. J. Am. Chem. Soc. 1965 , 87 , 4396-4397.

122. Frey, H. M.; Stevens, I. D. R. Proc. Chem. Soc. 1962 , 79-80.

123. Ganzer, G. A.; Sheridan, R. S.; Liu, M. T. H. J. Am. Chem. Soc. 1986 , 108 , 1517- 1520. 255 124. Soundararajan, N.; Platz, M. S.; Jackson, J. E.; Doyle, M. P.; Oon, S. M.; Liu, M. T. H.; Anand, S. M. J. Am. Chem. Soc. 1988 , 110 , 7143-7152.

125. Amrich, M. J.; Bell, J. A. J. Am. Chem. Soc. 1964 , 86 , 292-293.

126. Rosenberg, M. G.; Kam, S. M.; Brinker, U. H. Tetrahedron Lett. 1996 , 37 , 3235- 3238.

127. Moss, R. A.; Chang, M. J. Tetrahedron Lett. 1981 , 22 , 3749-3752.

128. Smith, R. A. G.; Knowles, J. R. J. Chem. Soc., Perkin Trans. 2 1975 , 686-694.

129. Bayley, H.; Knowles, J. R. Biochemistry 1978 , 17 , 2420-2423.

130. Frey, H. M.; Stevens, I. D. R. J. Chem. Soc. 1963 , 3514-3519.

131. Zhang, Y.; Kubicki, J.; Wang, J.; Platz, M. S. J. Phys. Chem. A 2008 , accepted for publication .

132. Schrader, T.; Sieg, A.; Koller, F.; Schreier, W.; An, Q.; Zinth, W.; Gilch, P. Chem. Phys. Lett. 2004 , 392 , 358-364.

133. Cohen, A. D.; Showalter, B. M.; Toscano, J. P. Org. Lett. 2004 , 6, 401-403.

134. Reichardt, C.; Schroeder, J.; Voehringer, P.; Schwarzer, D. Phys. Chem. Chem. Phys. 2008 , 10 , 1662-1668.

135. Kubicki, J.; Zhang, Y.; Wang, J.; Luk, H. L.; Peng, H.-L.; Vyas, S.; Platz, M. S. J. Am. Chem. Soc. 2009 , 131 , 4212-4213.

136. Senthilnathan, V. P.; Platz, M. S. J. Am. Chem. Soc. 1980 , 102 , 7637-7643.

137. Wong, M. W.; Wentrup, C. J. Org. Chem. 1996 , 61 , 7022-7029.

138. Cramer, C. J.; Dulles, F. J.; Falvey, D. E. J. Am. Chem. Soc. 1994 , 116 , 9787- 9788.

139. Dorigo, A. E.; Li, Y.; Houk, K. N. J. Am. Chem. Soc. 1989 , 111 , 6942-6948.

140. Roth, H. D. J. Amer. Chem. Soc. 1971 , 93 , 1527-1529.

141. Joines, R. C.; Turner, A. B.; Jones, W. M. J. Am. Chem. Soc 1969 , 91 , 7754-7755.

142. West, P. R.; Chapman, O. L.; LeRoux, J. P. J. Am. Chem. Soc. 1982 , 104 , 1779- 1782.

256 143. McMahon, R. J.; Abelt, C. J.; Chapman, O. L.; Johnson, J. W.; Kreil, C. L.; LeRoux, J. P.; Mooring, A. M.; West, P. R. J. Am. Chem. Soc. 1987 , 109 , 2456- 2469.

144. McMahon, R. J.; Chapman, O. L. J. Am. Chem. Soc. 1987 , 109 , 683-692.

145. Chapman, O. L.; McMahon, R. J.; West, P. R. J. Am. Chem. Soc. 1984 , 106 , 7973-7974.

146. Warmuth, R.; Marvel, M. A. Chemistry--A European Journal 2001 , 7, 1209-1220.

147. Jackson, J. E.; Soundararajan, N.; White, W.; Liu, M. T. H.; Bonneau, R.; Platz, M. S. J. Am. Chem. Soc. 1989 , 111 , 6874-6875.

148. Wang, J.; Kubicki, J.; Peng, H.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 6604- 6609.

149. Wang, J., Ph.D. Thesis, the Ohio State University, 2007.

150. Schmitz, E.; Ohme, R. Chem. Ber. 1962 , 95 , 795-802.

151. Frey, H. M.; Stevens, I. D. R. J. Chem. Soc. 1963 , 3514-1519.

152. Burdzinski, G. T.; Gustafson, T. L.; Hackett, J. C.; Hadad, C. M.; Platz, M. S. J. Am. Chem. Soc. 2005 , 127 , 13764-13765.

153. White, W. R., III; Platz, M. S. J. Org. Chem. 1992 , 57 , 2841-2846.

154. Huang, H.; Platz, M. S. J. Am. Chem. Soc. 1998 , 120 , 5990-5999.

155. Platz, M. S. Advances in Carbene Chemistry 1998 , 2, 133-174.

156. Burdzinski, G.; Rehault, J.; Wang, J.; Platz, M. S. J. Phys. Chem. A 2008 , 112 , 10108-10112.

157. Burdzinski, G. T.; Wang, J.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 3746-3747.

158. Wang, J.; Burdzinski, G.; Kubicki, J.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 2008 , 130 , 5418-5419.

159. Zhang, Y.; Kubicki, J.; Wang, J.; Platz, M. S. J. Phys. Chem. A 2008 , 112 , 11093- 11098.

160. Stoutland, P. O.; Dyer, R. B.; Woodruff, W. H. Science 1992 , 257 , 1913-1917.

161. Garavelli, M.; Negri, F.; Olivucci, M. J. Am. Chem. Soc. 1999 , 121 , 1023-1029. 257 162. Zgierski, M. Z.; Lim, E. C. J. Chem. Phys. 2005 , 122 , 111103.

163. Wolf, M. M. N.; Schumann, C.; Gross, R.; Domratcheva, T.; Diller, R. J. Phys. Chem. B 2008 , 112 , 13424-13432.

164. Brunner, J.; Senn, H.; Richards, F. M. J. Bio. Chem. 1980 , 255 , 3313-3318.

165. Richard A. G. Smith, J. R. K. J. Am. Chem. Soc. 1973 , 95 , 2.

166. Nibbering, E. T. J.; Elsaesser, T. Appl. Phys. B 2000 , 71 , 439-441.

167. Scott, A. P.; Platz, M. S.; Radom, L. J. Am. Chem. Soc. 2001 , 123 , 6069-6076.

168. Liu, M. T. H. Chem. Soc. Rev. 1982 , 11 , 127-140.

169. Buterbaugh, J. S.; Toscano, J. P.; Weaver, W. L.; Gord, J. R.; Hadad, C. M.; Gustafson, T. L.; Platz, M. S. J. Am. Chem. Soc. 1997 , 119 , 3580-3591.

170. Irikura, K. K.; Goddard, W. A., III; Beauchamp, J. L. J. Am. Chem. Soc. 1992 , 114 , 48-51.

171. Garcia, V. M.; Castell, O.; Reguero, M.; Caballol, R. Mol. Phys. 1996 , 87 , 1395- 1404.

172. Zhang, Y.; Burdzinski, G.; Kubicki, J.; Platz, M. S. Abstracts of Papers, 237th ACS National Meeting, Salt Lake City, UT, United States, March 22-26, 2009 2009 , ORGN-248.

173. Zuev, P. S.; Sheridan, R. S. J. Org. Chem. 1994 , 59 , 2267-2269.

174. Pliego, J. R., Jr.; De Almeida, W. B.; Celebi, S.; Zhu, Z.; Platz, M. S. J. Phys. Chem. A 1999 , 103 , 7481-7486.

175. Rabinovitch, B. S.; Rynbrandt, J. D. J. Phys. Chem. 1971 , 75 , 2164-2171.

176. Trenwith, A. B.; Rabinovitch, B. S. J. Chem. Phys. 1986 , 85 , 1694-1695.

177. Schwarzer, D.; Hanisch, C.; Kutne, P.; Troe, J. J. Phys. Chem. A 2002 , 106 , 8019- 8028.

178. Moss, R. A.; Terpinski, J.; Cox, D. P.; Denny, D. Z.; Krogh-Jespersen, K. J. Am. Chem. Soc. 1985 , 107 , 2743-2748.

179. Kazmirowski, H. G.; Landmann, H.; wALSMANN, P. Pharmazie 1969 , 24 , 378- 380.

180. Harriman, A.; Hosie, R. J. J. Chem. Soc., Faraday Trans. 2 1981 , 77 , 1695-1702. 258 181. Wagner, P. J.; Truman, R. J.; Scaiano, J. C. J. Am. Chem. Soc. 1985 , 107 , 7093- 7097.

182. Sakamoto, M.; Cai, X.; Hara, M.; Fujitsuka, M.; Majima, T. J. Phys. Chem. A 2005 , 109 , 4657-4661.

183. Turro, N. J.; Ramamurthy, V.; Cherry, W.; Farneth, W. Chem. Rev. 1978 , 78 , 125-145.

184. Kasha, M. Discuss. Faraday Soc. 1950 , No. 9 , 14-19.

185. Bogdanova, A.; Popik, V. V. J. Am. Chem. Soc. 2003 , 125 , 1456-1457.

186. Bogdanova, A.; Popik, V. V. J. Am. Chem. Soc. 2003 , 125 , 14153-14162.

187. Charvat, A.; Abmann, J.; Abel, B.; Schwarzer, D.; Henning, K.; Luther, K.; Troe, J. Phys. Chem. Chem. Phys. 2001 , 3, 2230-2240.

188. Platz, M. S. Kinetics and Spectroscopy of Carbenes and Biradicals ; Plenum: New York, 1990.

189. Jones, M., Jr.; Rettig, K. R. J. Am. Chem. Soc. 1965 , 87 , 4015-4016.

190. Closs, G. L.; Closs, L. E. J. Am. Chem. Soc. 1961 , 83 , 2015-2016.

191. Schreiner, P. R.; Karney, W. L.; Schleyer, P. v. R.; Borden, W. T.; Hamilton, T. P.; Schaefer, H. F., III. J. Org. Chem. 1996 , 61 , 7030-7039.

192. Hutton, R. S.; Manion, M. L.; Roth, H. D.; Wasserman, E. J. Amer. Chem. Soc. 1974 , 96 , 4680-4682.

193. Senthilnathan, V. P.; Platz, M. S. J. Am. Chem. Soc. 1981 , 103 , 5503-5511.

194. Cochran, E. L.; Adrian, F. J.; Bowers, V. A. J. Chem. Phys. 1964 , 40 , 213-220.

195. Closs, G. L.; Kaplan, L. R.; Bendall, V. I. J. Am. Chem. Soc. 1967 , 89 , 3376-3377.

196. Closs, G. L.; Boell, W. A.; Heyn, H.; Dev, V. J. Am. Chem. Soc. 1968 , 90 , 173- 178.

197. York, E. J.; Dittmar, W.; Stevenson, J. R.; Bergman, R. G. J. Am. Chem. Soc. 1973 , 95 , 5680-5687.

198. Pincock, J. A.; Morchat, R.; Arnold, D. R. J. Amer. Chem. Soc. 1973 , 95 , 7536- 7538.

259 199. Poutsma, J. C.; Nash, J. J.; Paulino, J. A.; Squires, R. R. J. Am. Chem. Soc. 1997 , 119 , 4686-4697.

200. Yoshimine, M.; Pacansky, J.; Honjou, N. J. Am. Chem. Soc. 1989 , 111 , 2785- 2798.

201. Honjou, N.; Pacansky, J.; Yoshimine, M. J. Am. Chem. Soc. 1985 , 107 , 5332- 5341.

202. Davis, J. H.; Goddard, W. A., III; Bergman, R. G. J. Am. Chem. Soc. 1977 , 99 , 2427-2434.

203. Freeman, P. K. J. Org. Chem. 2009 , 74 , 830-833.

204. Zuev, P. S.; Sheridan, R. S. J. Am. Chem. Soc. 2004 , 126 , 12220-12221.

205. Moss, R. A.; Tian, J.; Sauers, R. R.; Sheridan, R. S.; Bhakta, A.; Zuev, P. S. Org. Lett. 2005 , 7, 4645-4648.

206. Brewbaker, J. L.; Hart, H. J. Amer. Chem. Soc. 1969 , 91 , 711-715.

207. Davies, H. M. L.; Jin, Q. Proc. Natl. Acad. Sci. U. S. A. 2004 , 101 , 5472-5475.

208. Lovely, C. J.; Flores, J. A.; Meng, X.; Dias, H. V. R. Synlett 2009 , 129-132.

209. Doyle, M. P.; Yan, M.; Hu, W.; Gronenberg, L. S. J. Am. Chem. Soc. 2003 , 125 , 4692-4693.

210. Zhang, Y.; Kubicki, J.; Wang, J.; Platz, M. S. J. Phys. Chem. A 2008 , 112 , 11093- 11098.

211. Schmitz, C. Ph.D. Thesis, University of Bochum. 1986 .

212. Zhang, Y.; Burdzinski, G.; Kubicki, J.; Platz, M. S. J. Am. Chem. Soc. 2009 , 131 , 9646-9647..

213. Schwarzer, D.; Troe, J.; Votsmeier, M.; Zerezke, M. J. Chem. Phys. 1996 , 105 , 3121-3131.

214. Padwa, A.; Kennedy, G. D. J. Org. Chem. 1984 , 49 , 4344-4352.

215. Minato, H. Bull. Chem. Soc. Jpn. 1963 , 36 , 1020-1025.

216. Hess, G. C.; Kohler, B.; Likhotvorik, I.; Peon, J.; Platz, M. S. J. Am. Chem. Soc. 2000 , 122 , 8087-8088.

217. Wang, Y.; Hadad, C. M.; Toscano, J. P. J. Am. Chem. Soc. 2002 , 124 , 1761-1767. 260 218. Wang, Y.; Yuzawa, T.; Hamaguchi, H.-o.; Toscano, J. P. J. Am. Chem. Soc. 1999 , 121 , 2875-2882.

219. Maier, G.; Senger, S. Angew. Chem., Int. Ed. Engl. 1994 , 33 , 558-559.

220. Maier, G.; Lautz, C.; Senger, S. Chem. Eur. J. 2000 , 6, 1467-1473.

221. Albers, R.; Sander, W.; Ottosson, C.-H.; Cremer, D. Chem. Eur. J. 1996 , 2, 967- 973.

222. Arndt, F. Org. Synth. 1935 , 15 , 3.

223. Frey, H. M. Pure and Applied Chemistry 1964 , 9, 527-537.

224. Friedman, L.; Shechter, H. J. Am. Chem. Soc. 1959 , 81 , 5512-5513.

225. Mansoor, A. M.; Stevens, I. D. R. Tetrahedron Lett. 1966 , 1733-1737.

226. Closs, G. L. J. Am. Chem. Soc. 1962 , 84 , 809-813.

227. Friedman, L.; Shechter, H. J. Am. Chem. Soc. 1961 , 83 , 3159-3160.

228. Kirmse, W.; Buecking, H. W.; Buecking, H. W. Tetrahedron Lett. 1967 , 1833- 1835.

229. Rando, R. R. J. Am. Chem. Soc. 1970 , 92 , 6706-6707.

230. Rando, R. R. J. Am. Chem. Soc. 1972 , 94 , 1629-1631.

231. Corey, E. J.; Felix, A. M. J. Am. Chem. Soc. 1965 , 87 , 2518-2519.

232. Candeias, N. R.; Gois, P. M. P.; Veiros, L. F.; Afonso, C. A. M. J. Org. Chem. 2008 , 73 , 5926-5932.

233. Tomioka, H.; Kitagawa, H.; Izawa, Y. J. Org. Chem. 1979 , 44 , 3072-3075.

234. Tomioka, H.; Kondo, M.; Izawa, Y. J. Org. Chem. 1981 , 46 , 1090-1094.

235. Kaplan, F.; Meloy, G. K. J. Am. Chem. Soc. 1966 , 88 , 950-956.

236. Chaimovich, H.; Vaughan, R. J.; Westheimer, F. H. J. Am. Chem. Soc. 1968 , 90 , 4088-4093.

237. Hamm, P.; Ohline, S. M.; Zinth, W. J. Chem. Phys. 1997 , 106 , 519-529.

238. Mennucci, B.; Tomasi, J. J. Chem. Phys. 1997 , 106 , 5151-5158.

261 239. Peng, C.; Ayala, P.; Schlegel, H. B.; Frisch, M. J. J. Comput. Chem. 1996 , 17 , 49- 56.

240. Shafer, J.; Barnowsky, P.; Laursen, R.; Finn, F.; Westheimer, F. H. J. Biol. Chem. 1966 , 241 , 421-427.

262 APPENDIX:

1H and 13 C NMR SPECTRA

263 H NMRH 1 H H 2 CN 5 H 6 C

264 C NMR 3 1 H H 2 CN 5 H 6 C

265 H NMRH 1

3 CH 2 CN 5 H 6 C

266 C NMR 3 1

3 CH 2 CN 5 H 6 C

267 H NMR NMR H 1

3 CH 2 CH 2 CN 5 H 6 C

268 C NMR 3 1

3 CH 2 CH 2 CN 5 H 6 C

269 H NMRH 1 -Bu t - 2 CN 5 H 6 C

270 C NMR 3 1 -Bu t - 2 CN 5 H 6 C

271 H NMRH 1

3 CD 2 CN 5 H 6 C

272

H NMR H 1

3 CH 2 CN 4 H 6 OC 3 -CH p

273 C NMR 3 1

3 CH 2 CN 4 H 6 OC 3 -CH p

274 H NMRH 1 H H 2 -BpCN p

275 C NMR 3 1 H H 2 -BpCN p

276

277

278

279

280

H NMRH 1

3 CH 2 CO 2 PhCH=CHCN

281 C NMR 3 1

3 CH 2 CO 2 PhCH=CHCN

282