Modern Physical Organic Chemistry

Eric V. Anslyn

Dennis A. Dougherty

University Science Books
www. u

Uni versity Science Books

Production Manager: Christine Taylor
Manuscript Editor: John Murdzek
Designer: Robert Ish i
Illustrator: Lineworks
Compositor: Wilsted & Taylor Publishing Services
Printe r & Binder: Edwards Brothers, Inc.

This book is p rinted on acid-free paper.

Cop yright © 2006 by University Science Books

Reprodu ction or transla tion of any part o f this work beyond tha t permitted by
Section 107 or 108 of the 1976 Un ited Sta tes Copyright Act w ithout the permission
of the copyrigh t owner is unlawful. Requests for permission or furth er informatio n
shou ld be addressed to the Permissions Department, University Science Books.

Library o f Congress Cataloging-in-Publica tion Data

An slyn, Eric V., 1960-
Modern physical organic chemistry I Eric V. Anslyn, Dennis A. Dougherty.
p. em .
In cludes bibliographi cal references and index.
ISBN 978-1-891389-31-3 (alk. paper)
1. C hem istry, Ph ysica l organic. I. Dou gher ty, Dennis A., 1952- II. Title.

QD476.A57 2004
547' .13-d c22

Printed in the United States of America
10 9 8 7 6 5 4

Abbreviated Contents

PART I: Molecular Structure and Thermodynamics
CHAPTER 1. Introduction to Structure and Models of Bonding 3
2. Strain and Stability 65
3. Solutions and Non-Covalent Binding Forces 145
4. Molecular Recognition and Supramolecular Chemistry 207
5. Acid-Base Chemistry 259
6. Stereochemistry 297

PART II: Reactivity, Kinetics, and Mechanisms
CHAPTER 7. Energy Surfaces and Kinetic Analyses 355
8. Experiments Related to Thermodynamics and Kinetics 421
9. Catalysis 489
10. Organic Reaction Mechanisms, Part 1:
Reactions Involving Additions and / or Eliminations 537
11. Organic Reaction Mechanisms, Part 2:
Substitutions at Aliphatic Centers and Thermal
Isomerizations / Rearrangements 627
12. Organotransition Metal Reaction Mechanisms and Catalysis 705
13. Organic Polymer and Materials Chemistry 753

PART III: Electronic Structure: Theory and Applications

CHAPTER 14. Advanced Concepts in Electronic Structure Theory 807
15. Thermal Pericyclic Reactions 877
16. Photochemistry 935
17. Electronic Organic Materials 1001

APPENDIX 1. Conversion Factors and Other Useful Data 1047
2. Electrostatic Potential Surfaces for Representative Organic Molecules 1049
3. Group Orbitals of Common Functional Groups:
Representative Examples Using Simple Molecules 1051
4. The Organic Structures of Biology 1057
5. Pushing Electrons 1061
6. Reaction Mechanism Nomenclature 1075

INDEX 1079



List of Highlights xix 1.3.2 Using Grou p Orbitals to Make Ethy lene 38
Preface xxiii 1.3.3 The Effects of Heteroatoms-Formaldehyde 40
Acknowledgments xxv 1.3.4 Making More Complex Alkanes 43
A Note to the Instructor xxvii 1.3.5 Three More Examples of Building Larger
Molecules from Group Orbitals 43
Propene 43
PART I Methyl Chloride 45
THERMODYNAMICS 1.3.6 Group Orbitals of Representative TI Systems:
Benzene, Benzyl, and Allyl 46
CHAPTER 1: Introduction toStructure and 1.3.7 Understanding Common Functional
Groups as Perturbations of Allyl 49
Models of Bonding 3
1.3.8 The Three Center-Two Electron Bo nd 50
Intent and Purpose 3 1.3.9 Summary of the Concepts Involved in
Our Second Model of Bonding 51
1.1 A Review of Basic Bonding Concepts 4
1.1.1 Quantum Numbers and Atomic Orbitals 4 1.4 Bonding and Structures of Reactive Intermediates 52
1.1.2 Electron Configurations and Electronic Diagrams 5 1.4.1 Carbocations 52
1.1.3 Lewis Structures 6 Carbenium Ions 53
1.1.4 Formal Charge 6 Interplay with Carbonium Ions 54
1.1.5 VSEPR 7 Carbonium Ions 55
1.1.6 Hybridization 8 1.4.2 Carbanions 56
1.1.7 A Hybrid Valence Bond / MolecularOrbital 1.4.3 Radicals 57
Model of Bonding 10 1.4.4 Carbenes 58
Creating Localized CJand n Bonds 11
1.1.8 Polar Covalent Bonding 12 1.5 A Very Quick Look at Organometallic
Electronegativity 12 and Inorganic Bonding 59
Electrostatic Potential Surfaces 14
Inductive Effects 15 Summary and Outlook 61
Group Eiectronegativities 16
Hybridization Effects 17 EXERCISES 62
1.1.9 Bond Dipoles, Molecular Dipoles, FURTHER READING 64
and Quadrupoles 17
Bond Dipoles 17
Molecular Dipole Moments 18
Molecular Quadrupole Moments 19
1.1.10 Resonance 20
1.1.11 Bond Lengths 22 CHAPTER 2: Strain and Stability 65
1.1.12 Polarizability 24
1.1.13 Summary of Concepts Used for the Simplest Intent and Purpose 65
Model of Bonding in Organic Structures 26 2.1 Thermochemistry of Stable Molecules 66
2.1.1 The Concepts of Internal Strain
1.2 A More Modern Theory of Organic Bonding 26 and Relative Stability 66
1.2.1 Molecular Orbital Theory 27 2.1.2 Types of Energy 68
1.2.2 A Method for QMOT 28 Gibbs Free Energy 68
1.2.3 Methyl in Detail 29 Enthalpy 69
Planar Methyl 29 Entropy 70
The Walsh Diagram: Pyramidal Methyl 31 2.1.3 Bond Dissociation Energies 70
"Group Orbitals"for Pyramidal Methyl 32 Using BDEs to Predict Exothermicity
Putting the Electrons In - The MH3 System 33 and Endothermicity 72
1.2.4 The CH2 Group in Detail 33 2.1.4 An Introduction to Potential Functions
The Walsh Diagram and Group Orbitals 33 and Surfaces- Bond Stretches 73
Putting the Electrons In- The MH 2 System 33 Infrared Spectroscopy 77
2.1.5 Heats of Formation and Combustion 77
1.3 Orbital Mixing-Building Larger Molecules 35
2.1.6 The Group Increment Method 79
1.3.1 Using Group Orbitals to Make Ethane 36
2.1.7 Strain Energy 82 Vll


2.2 Thermoch emistry of Reactive Interm ediates 82 Electrostatic Interactions 131
2.2.1 Stability vs. Persistence 82 Hydrogen Bonding 131
2.2.2 Radicals 83 The Parameterization 132
BDEs as a Measure of Stability 83 Heat of Formation and Strain Energy 132
Radical Persistence 84 2.6.2 General Commen ts on the Molecular
Group Increments for Radicals 86 Mechanics Method 133
2.2.3 Carbocations 87 2.6.3 Molecul ar Mech anics on Biomolecules and
Hydride Ion Affinities as a Measure of Stability 87 Unnatu ral Polymers-"Modeling" 135
Lifetimes ofCarbocations 90 2.6.4 Molecu lar Mech anics Studies of Reactions 136
2.2.4 Carbanions 91
2.2.5 Summary 91 Summary and Outlook 137

2.3 Rela tionships Between Struc ture and Energe tics- EXERCISES 138
Basic Conformationa l Analysis 92 FURT HER READI NG 143
2.3.1 Acyclic Systems-Torsiona l Poten tial Surfaces 92
Ethane 92
Butane- The Gauche Interaction 95
Barrier Height 97
Barrier Foldedness 97
Tetraalky!ethanes 98 CH A PT ER 3: Solu tions
and Non-Covalent
The g+g- Pentane Interaction 99 Binding Forces 145
Allylic(A 1•3) Strain 100 Intent and Purpose 145
2.3.2 Basic Cyclic Systems 100
Cyclopropane 100 3.1 Solvent and Solution Properties 145
Cyc!obutane 100 3.1.1 N ature Abh ors a Vacuum 146
Cyc!opentalle 101 3.1.2 Solvent Scales 146
Cyc!ohcxanc 102 Dielectric Constant 147
Larger Rings- Transamwlar Effects 107 Other Solvent Scales 148
Group ln creme11t Corrections for Ring Systems 109 Heat of Vaporization 150
Ri11g Torsional Modes 109 Surface Tension and Wetting 150
Bicyc!ic Ring Systems 110 Water 151
Cycloalkencs and Bredt's Rule 110 3 .1.3 Solubility 153
SuJmJtan; of Conformational Analysis and General Overview 153
Its Connection to Strain 112 Shape 154
Using the "Like-Dissolves-Like" Paradigm 154
2.4 Electronic Effects 112 3.1.4 Solute Mobility 155
2.4.1 Inte ractions Involving TI Systems 112 Diffusion 155
Subs titution 011 Alkenes 112 Fick's Law of Diffusion 156
Confor/1/atiolls of Substituted Alkenes 113 Correlation Times 156
Conjugation 115 3.1.5 The Thermodynamics of Solutions 157
Aromaticity 116 Chemical Potential 158
Antiaromaticity, An Unu sual Des tabilizing Effect 117 The Thermodynamics of Reactions 160
NM R Chemical Shifts 118 Calculating t1H 0 and flS o 162
Polycyclic Aromatic Hydrocarbons 119
Large Annulenes 119 3.2 Binding Forces 162
2.4.2 Effects of Multiple Heteroatoms 120 3.2.1 Ion Pairing Interacti ons 163
Bond Le11gth Effects 120 Salt Bridges 164
Orbital Effects 120 3.2.2 Electrosta tic In teractions In volving Dipoles 165
Ion-Dipole Interactions 165
2.5 Highly-Strai ned Molecul es 124 A Simple Model of Tonic Solvation -
2.5 .1 Long Bonds and Large Angles 124 The Born Equation 166
2.5.2 Sma ll Rings 125 Dipole-Dipole Interactions 168
2.5.3 Very Large Rotation Barriers 127 3.2.3 Hydrogen Bon d ing 168
Geometries 169
2.6 Molecular Mechanics 128 Strengths of Normal Hydrogen Bonds 171
2.6.1 The Molecular Mechanics Mod el 129 i. Solvation Effects 171
Bond Stretching 129 ii. Electronegativity Effects 172
Angle Bending 130 iii. Resonance Assisted Hydrogen Bonds 173
Torsion 130 iv. Polarization Enhanced Hydrogen Bonds 174
Nonbonded Interactions 130 v. Secondary Interactions in Hydrogen
Cross Terms 131 Bonding Systems 175


vi. CoopemtivihJ in Hydrogen Bonds 175 Molecular Recognition via Hydrogen
Vibra tional Properties of Hydrogen Bonds 176 Bonding in Water 232
Short-Strong Hydrogen Bonds 177 4.2.4 Molecular Recognition with a Large
3.2.4 '1T Effects 180 H ydrop hobic Component 234
Cation-rc Interactions 181 Cyclodextrins 234
Polar-rc Interactions 183 Cyclophanes 234
Aromatic-Aromatic Interactions (rc Stacking) 184 A Sum mary of the Hydrophobic Co mpoueut
The Arene-Perfluoroarene Interaction 184 of Molecular Recognition in Water 238
rr: Donor-Acceptor Interactions 186 4.2.5 Molecul ar Recognition with a La rge '1T
3.2.5 Indu ced -D ipole Interactio ns 186 Componen t 239
/on- Induced-Dipole Interactions 187 Ca tion-lf Interactions 239
Dipole-Induced-Dipole Interactions 187 Polar-lf and Related Effects 241
lndttced-Dipole-lnduced-Dipole Interactions 188 4.2.6 Summary 241
Sunnttarizing Monopole, Dipole, and
Induced-Dipole Binding Forces 188 4.3 Supramolecul ar Chemistry 243
3.2.6 The Hyd ro phobic Effect 189 4.3. 1 Supra molecular Assembl y of Complex
Aggregation of Orga nics 189 Architectures 244
The Origin of the Hydro phobic Effect 192 Self-Assembly via Coordination Compounds 244
Self-Assernbly via Hydrogen Bonding 245
3.3 Computational Mod eling of Solva tio n 194 4.3.2 Novel Supra molecul a r Architectu res-Cate nanes,
3.3.1 Conti nuum Solva ti on Mod els 196 Rotaxa nes, and Knots 246
3.3.2 Expli cit Solva tionModel s 197 Nano technology 248
3.3.3 Monte Carlo (MC) Method s 198 4.3.3 Container Compo w1ds-Molecu les w ithin
3.3.4 Molecula r Dyn amics (MD) 199 M olecules 249
3.3.5 Stati sti cal Perturbation Theory/
Free Energy Pe rturba tion 200 Summary and Outlook 252

Sum mary and Outlook 201 EXERCISES 253

CHAPTER 4: M olecular Recognition and CHAPTERS: Acid-Base Chemistry 259
Supramolecular Chemistry 207
Intent a nd Purpose 259
In tent and Purpose 207
5.1 Bro ns ted Acid-Base Chemistry 259
4.1 Th erm od ynamic Anal yses of Binding
Phen omena 207 5.2 Aqueous Solutions 261
4.1.1 General Thermodynamics of Bind ing 208 5.2.1 pK. 261
The Relevance of the Standard State 210 5.2.2 pH 262
The Influence of a Change in Hea t Capacity 212 5.2.3 The Leveli ng Effect 264
Coopera tivity 213 5.2.4 Acti vity vs. Concentration 266
En thalpy-En tropy Compensation 216 5.2.5 Acidi ty Fu nction s: Acidity Sca les fo r Hig hl y
4.1.2 The Binding Isotherm 21 6 Con centrated Acidic Solutions 266
4.1.3 Ex perimental Methods 21 9 5.2.6 Super Acid s 270
U V/Vis or Fluorescence Methods 220
NMRMethods 220 5.3 Nonaqueous Sys tems 271
Isothermal Calorimetry 221 5.3.1 p K, Shi fts at Enzyme Acti ve Sites 273
5.3.2 Solu tion Phase vs. C as Ph ase 273
4.2 M ol ecular Recognition 222
4.2.1 Comple mentarity and Preorgan ization 224 5.4 Predicting Acid Stre n gth in Solution 276
Crowns, Cryptands, and Sphera nds -Molecular 5.4. 1 Me thods Used to Measure Wea k Acid Strength 276
Recogn ition with a Large Ion-Dipole Component 224 5.4.2 Tw o Gu iding Princip les for Predi ctin g
T·weezers and Clefts 228 Rela ti ve Acidities 277
4.2.2 Molecul ar Recognition w ith a La rge 5.4.3 Electronega tivity and Inducti on 278
Ion Pairing Component 228 5.4.4 Reson an ce 278
4.2.3 Mo lecul ar Recognition with a Large H ydrogen 5.4.5 Bo nd Stre ngth s 283
Bonding Componen t 230 5.4.6 Electrostatic Effects 283
Represen tative Structures 230 5.4.7 H ybridi za ti on 283


5.4.8 Aromaticity 284 6.6.3 NonplanarGraphs 326
5.4.9 Solvation 284 6.6.4 Achievements in Top ologica l and Supram olecular
5.4.10 Cationic Organic Structu res 285 Stereochemistry 327

5.5 Acids and Bases of Biological Interest 285 6.7 Stereochemical Issues in Polymer Chemistry 331

5.6 Lewis Acids/Bases and Electrophiles/ 6.8 Stereochemical Issues in Chemical Biology 333
Nucleophiles 288 6.8.1 Th e Linkages of Proteins, N ucleic Acids,
5.6.1 The Concept of Hard and Soft Acids and Bases, Genera l and Polysacch arides 333
Lessons for Lewis Acid-Base In teractions, and Relative Proteins 333
Nucleophilicity and Electrophilicity 289 Nucleic Acids 334
Polysaccharides 334
Summary and O utlook 292 6.8.2 Helicity 336
EXERCISES 292 Syn thetic Helical Polymers 337
FURTHER READING 294 6.8.3 Th e Origin of Ch ira lity in Nature 339

6.9 Stereochemical Terminology 340

CHAPTER 6: Stereochemistry 297 Summary and Outlook 344
Intent and Pu rpose 297 EXER C ISES 344
6.1 Stereogenicity and Stereoisomerism 297
6.1.1 Basic Concepts and Term inology 298
Classic Terminology 299
More Modem Terminology 301
6.1.2 Stereochemical Descriptors 303
R,S System 304
o and L 304 MECHANISMS
Erythro and Tlneo 305
Helical Descriptors- M and P 305 CHAPTER 7: Energy Surfaces and
Ent nnd Epi 306 Kinetic Analyses 355
Using Descriptors to Compare Structures 306
6.1.3 Di stin gu ishing Enan tiomers 306 Intent and Purpose 355
Optical Activity nnd Chirality 309
7.1 Energy Surfaces and Related Concepts 356
Why is Plane Polarized Light Rotated
7.1.1 Energy Surfaces 357
by a Chirnl Medium? 309
7.1.2 Reaction Coordinate Diagrams 359
Circular Dichroism 310
7.1.3 What is the Nature of the Activa ted
X-Ray Crystallography 310
Complex/ Transition State? 362
6.2 Sym metry and Stereochemis try 311 7.1.4 Rates and Rate Constants 363
6.2 .1 Basic Symmetry Ope rations 311 7.1.5 Reaction Order and Rate Laws 364
6.2 .2 Chirality and Symmetry 311
6.2.3 Symmetry Arguments 313 7.2 Transition State Theory (TST) and Related Topics 365
6.2 .4 Focusing on Carbon 314 7.2.1 The Mathem ati cs of Transition State Theory 365
7.2.2 Relationship to the Arrhenius Rate Law 367
6.3 Top icity Relations h ips 315 7.2.3 Boltzmann Distributions and Temp erature
6.3.1 Homotopic, Enantiotopic, and Diastereotopic 315 Dependence 368
6.3.2 To pi city Descri ptors-Pro-R I Pro-S and Re I Si 316 7.2.4 Revisi ting "Wh at is the Na tu re of the Acti va ted
6.3.3 Chirotopicity 317 Complex?" and Why Does TST Work? 369
7.2.5 Experimental Determinations of Activa tion Param eters
6.4 Reac tion Stereochemis try: Stereoselectivity and Arrhenius Parameters 370
and Stereospecifi city 317 7.2.6 Examples of Acti va tion Param eters and
6.4.1 Simple Guidelines for Reaction Stereoch emis try 317 Their Interpretations 372
6.4.2 Stereospecific and Stereoselective Reactions 319 7.2.7 Is TST Com p letely Correct? The Dynamic Beh avior
of Organic Reactive Intermedia tes 372
6.5 Symmetry and Time Scale 322
7.3 Postulates and Principles Related
6.6 Topological and Supramolecular Stereochemistry 324 to Kinetic Analysis 374
6.6.1 Loops and Kno ts 325 7.3.1 The Hammond Postulate 374
6.6.2 Topological Chirality 326 7.3.2 The Reacti vity vs. Selectivity Principle 377


7.3.3 The Curtin-Hammett Principle 378 Primary Kinetic Isotope Effects for Linear
7.3.4 Microscopic Reversibility 379 Transition States as a Function ofExothermicity
7.3.5 Kinetic vs. Thermodynamic Control 380 and Endothermicity 425
Isotope Effects for Linear vs. Non-Linear
7.4 Kinetic Experiments 382 Transition States 428
7.4.1 How Kinetic Experimen ts are Performed 382 8.1 .3 The Origin of Secondary Kineti c Isotope Effects 428
7.4.2 Kinetic Analyses for Simple Mechan isms 384 Hybridization Changes 429
First Order Kinetics 385 Steric isotope Effects 430
Second Order Kinetics 386 8.1.4 Equilibrium Isotope Effects 432
Pseudo-First Order Kinetics 387 Isotopic Perturbation of Eqtlilibrium-
Equilibriu111 Kinetics 388 Applications to Carbocations 432
Initial-Rate Kinetics 389 8.1.5 Tunne ling 435
Tal111lating a Series ofConwton Kinetic Scenarios 389 8.1.6 Solve nt Isotope Effects 437
Fractionation Factors 437
7.5 Complex Reactions-Deciphering Mechanisms 390 Proton In ventories 438
7.5.1 Steady Sta te Kinetics 390 8.1.7 Heavy Atom Isotope Effects 441
7.5.2 Using the SSA to Predict Changes 8.1.8 Summ ary 441
in Kinetic Order 395
7.5.3 Saturation Kinetics 396 8.2 Substituent Effects 441
7.5.4 Prior Rap id Equilibria 397 8.2.1 The Origin of Substituent Effects 443
Field Effects 443
7.6 Methods for Following Kinetics 397 Indu ctive Effects 443
7.6.1 Reactions w ith Half-Lives Greater Resonance Effects 444
than a Few Seconds 398 Polarizability Effects 444
7.6.2 Fast Kinetics Techniques 398 Steric Effects 445
Flow Techniques 399 Solvation Effects 445
Flash Photolysis 399
Pnlse Radio/ ysis 401
8.3 Hammett Plots-The Most Common LFER.
7.6.3 Re laxation Methods 401
A General Method for Examining Changes
7.6.4 Summary of Kinetic Analyses 402
in Charges During a Reaction 445
8.3.1 Sigma (cr) 445
7.7 Calculating Rate Constants 403
8.3.2 Rho (p) 447
7.7.1 Marcus Theory 403
8.3.3 The Power of Hamme tt Plots for
7.7.2 Marcus Theory Ap plied to Electron Transfer 405
Deciphering Mechanisms 448
8.3.4 Dev iati on s from Linearity 449
7.8 Considering Multiple Reaction Coordinates 407
8.3.5 Separa ting Resonance from Induction 451
7.8.1 Variation in Tra nsition Sta te Stru ctures Across
a Series of Re la ted Reactions-An Example
Using Substitution Reactions 407 8.4 Other Linear Free Energy Relationships 454
7.8.2 More O'Ferrall-Jencks Plots 409 8.4.1 Steric and Polar Effects-Taft Parameters 454
7.8.3 Changes in Vibrational State Along the Reaction 8.4.2 Solvent Effects- Grun wa ld- Winstein Plots 455
Coordinate-Relating the Third Coordinate 8.4.3 Schleyer Ad aptation 457
to Entropy 412 8.4.4 Nucleophilicity and N ucleofuga ljty 458
Basicity/Acidity 459
Summary and Outlook 413 Solvation 460
Polarizability, Basicity, and Solvationlnterplay 460
EXERC ISES 413 Shape 461
FURTHER READING 417 8.4.5 Swa in-Scott Parameters-Nucleophilicity
Parameters 461
8.4.6 Ed wards and Ritchie Correlati ons 463

8.5 Acid-Base Related Effects-
CHAPTER 8: Experiments Related to
Bnmsted Relationships 464
Thermodynamics and Kinetics 421
8.5.1 fJNu c 464
Intent and Purpose 421 8.5.2 f3Lc 464
8.5.3 Acid-Base Ca talysis 466
8.1 Isotope Effects 421
8.1.1 The Experiment 422 8.6 Why do Linear Free Energy Relationships Work? 466
8.1.2 The Origin of Primary Kinetic Isotope Effects 422 8.6.1 Gene ral Mathematics ofLFERs 467
Reaction Coordinate Diagrams and Isotope 8.6.2 Conditions to Create an LFER 468
Effects 424 8.6.3 The Isokine tic or Isoequilibrium Temperature 469


8.6.4 Wh y does Enthalpy-Entropy 9.3.4 Concerted or Sequential General-Acid-
Compensa tion Occur? 469 General-Base Catalysis 515
Steric Effects 470 9.3.5 The Bremsted Catalysis Law
Solvation 470 and Its Ramifications 516
A Linear Free Energy Relationship 516
8.7 Summary of Linear Free Energy Relationships 470 The Meaning of a and /3 517
a+/3=1 518
8.8 Miscellaneous Experiments for Deviations from Linearity 519
Studying Mechanisms 471 9.3.6 Predicting General-Add or
8.8.1 Productldentifi cation 472 General-Base Catalysis 520
8.8.2 Changing the Reactant Stru cture to Divert The Libido Rule 520
or Trap a Proposed Intermediate 473 Potential Energy Surfaces Dictate
8.8.3 Trapping and Competition Experiments 474 General or Specific Catalysis 521
8.8.4 Checking fo r a Common In termediate 475 9.3.7 The Dynamics of Proton Transfers 522
8.8.5 Cross-Over Experiments 476 Marcus Analysis 522
8.8.6 Stereochemical Analysis 476
8.8.7 Isotope Scrambling 477 9.4 Enzymatic Catalysis 523
8.8.8 Techniques to Stud y Radicals: Clocks and Traps 478 9.4.1 Michaelis-MentenKinetics 523
8.8.9 Direct Isolation and Characterization 9.4.2 The Meaning of KM, kcau and kcatf KM 524
of an Intermediate 480 9.4.3 Enzyme Active Sites 525
8.8.10 Transien t Spectroscopy 480 9.4.4 [S] vs. KM-Reaction Coordina te Diagrams 527
8.8.11 Stable Media 481 9.4.5 Sup ramolecular Interactions 529

Summary and Outlook 482 Summary and Outlook 530


CHAPTER 9: Catalysis 489
CHAPTER 10: Organic Reaction Mechanisms,
Intent and Purpose 489 Part 1: Reactions Involving Additions
and/or Eliminations 537
9.1 General Principles of Catalysis 490
9.1.1 Binding the Transition State Better Intent and Purpose 537
th an the Gro und State 491
10.1 Predicting Organic Reactivity 538
9.1.2 A Thermodynamic Cycle Ana lysis 493
10.1.1 A Useful Paradigm for Polar Reactions 539
9.1.3 A Spa tial Temporal Approach 494
Nucleophiles and Electrophiles 539
Lewis Acids and Lewis Bases 540
9.2 Forms of Catalysis 495
Donor-A cceptor Orbital Interactions 540
9.2.1 "Binding" is Akin to So lvation 495
10.1.2 Predicting Radical Reactivity 541
9.2.2 Proximity as a Binding Phenomenon 495
10.1.3 In Preparation for the Follow ing Sections 541
9.2.3 Electro philic Ca talysis 499
Electrostatic fnteractions 499 -ADDITION REACTIONS- 542
Me tal Jon Catalysis 500 10.2 Hydration of Carbonyl Structures 542
9.2.4 Acid-Base Cata lysis 502 10.2.1 Acid-Base Catalysis 543
9.2.5 Nucleophili c Catalysis 502 10.2.2 The Thermodynamics of the Formation
9.2.6 Cova lent Catalysis 504 of Geminal Diols and H emiacetals 544
9.2.7 Strain and Distortion 505
9.2.8 Phase Transfer Catalysis 507 10.3 Electrophilic Addition of Water to Alkenes
and Alkynes: Hydration 545
9.3 Brans ted Acid-Base Catalysis 507 10.3.1 Electron Pushing 546
9.3.1 SpecificCatal ysis 507 10.3.2 Acid-Catalyzed Aqueous Hydration 546
The Mathematics of Specific Catalysis 507 10.3.3 Regiochemistry 546
Kine tic Plots 510 10.3.4 Alkyne Hydrati on 547
9.3.2 General Catalysis 510
The Mathematics of General Catalysis 511 10.4 Electrophilic Addition of Hydrogen Halides
Kinetic Plots 512 to Alkenes and Alkynes 548
9.3.3 A Kinetic Equi valency 514 10.4.1 Electron Pushing 548

C O N TE TS Xlll

10.4.2 Experimental Observa tions Related to 10.12 .2 Ste reoch emical and Isotope
Regiochemistry and Stereochemistry 548 Labeling Evidence 577
10.4.3 Addition to Alky nes 551 10.12.3 Ca talysis of the H ydrolysi s of Acetals 578
10.12.4 Stereoelectronic Effects 579
10.5 Electrophilic Addition of Halogens 10.12.5 Cr0 3 Oxidation-The Jones Reagent 580
to Alkenes 551 Electron Pushing 580
10.5.1 Electron Pushing 551 A Few Experimental Observations 581
10.5.2 Stereochemistry 552
10.5.3 Other Ev idence Supporting a cr Complex 552 10.13 Elimination Reactions for Aliphatic Systems-
10.5.4 Mecha ni sti c Variants 553 Formation of Alkenes 581
10.5.5 Add iti on to Alkynes 554 10.13.1 Electron Pushing and Defi niti ons 581
10.13.2 Some Experimental Observations
10.6 Hydroboration 554 for E2 and E1 Reactions 582
10.6.1 ElectronPu shing 555 10.13.3 Contrasti n g Eli mina tion
10.6.2 Ex pe rimental Observa tions 555 and Subs titution 583
10.13.4 Anot he r Possibility- E1cB 584
10.7 Epoxidation 555 10.13.5 Kine ti cs a nd Experimental Observations
10.7.1 Elect ron Pushing 556 for E1cB 584
10.7.2 Ex perimental Observations 556 10.13.6 Contrasting E2, E1, and E1cB 586
10.13.7 Regiochemi stry of Elim inations 588
10.8 Nucleophilic Additions to 10.13.8 Stereochemistry of Elim inat ions-
Carbonyl Compounds 556 Orbital Considerations 590
10.8.1 Electron Pushing for a Few N ucleophilic 10.13.9 Dehydration 592
Additions 557 Electron Pushing 592
10.8.2 Experimenta l Ob serva ti ons for Other Mechanistic Possibilities 594
Cyanohydrin Formation 559 10.13.10 Therm al Eliminations 594
10.8.3 Experimental Observations for
Grignard Reactions 560 10.14 Eliminations from Radical Intermediates 596
10.8.4 Experime ntal Observations in
10.8.5 Orb ita l Considera tion s 561
The Biirgi-Ounitz Angle 561 10.15 The Addition of Nitrogen Nucleophiles
Orbital Mixing 562 to Carbonyl Structures, Followed
10.8.6 Conformational Effects in Addition s by Elimination 597
to Carbony l Compounds 562 10.15.1 Electron Pushing 598
10.8.7 Stereochemistry of ucleophilic Additions 563 10.15.2 Acid-Base Catalysis 598

10.9 Nucleophilic Additions to Olefins 567 10.16 The Addition of Carbon Nucleophiles,
10.9.1 Electron Pushing 567 Followed by Elimination-
10.9.2 Ex perim ental Observations 567 The Wittig Reaction 599
10.9.3 Regioche mistry of Addition 567 10.1 6.1 Electron Pushing 600
10.9.4 Baldwin's Rules 568
10.17 Acyl Transfers 600
10.10 Radical Additions to Unsaturated 10.17.1 Ge ne ral Electron-Pushing Schemes 600
Systems 569 10.17.2 Isotope Scrambling 601
10.10 .1 Electron Pu shin g for Radical Ad ditions 569 10.17.3 Predicting the Site of Cleavage for
10.10.2 Radi ca l Initi ators 570 Acy l Transfers from Esters 602
10.10.3 Chain Transfer vs. Polymeriza ti on 571 10.17.4 Ca talysis 602
10.10.4 Termination 571
10.10.5 Regiochemistry of Rad ical Additio ns 572 10.18 Electrophilic Aromatic Substitution 607
10.18.1 Electron Pushing fo r Electro phili c
10.11 Carbene Additions and Insertions 572 Aromatic Substituti ons 607
10.11 .1 Electron Pushing for Ca rbene Reactions 574 10.18.2 Kineti cs and Isotope Effects 608
10.11 .2 Carbene Genera tio n 574 10.18.3 Intermediate Complexes 608
10.11.3 Experimental Observations for 10.18.4 Regiochemistry and Relati ve Rates of
Carb ene Reactions 575 Aroma ti c Substituti on 609

- ELIMIN ATI ONS- 576 10.19 Nucleophilic Aromatic Substitution 611
10.12 Eliminations to Form Carbonyls or "Carbonyl-Like" 10.19.1 Electro n Pushing for Nucleophilic
Intermediates 577 Aroma ti c Substitution 611
10.12.1 Electron Pushing 577 10.19.2 Experimental Observa ti ons 611


10.20 Reactions Involving Benzyne 612 11 .5.9 Structure-Func tion Correlation s
10.20.1 Electron Pushing for Ben zyne Reactions 612 w ith the Nucleophile 648
10.20.2 Ex perimental Observa tions 613 11 .5.10 Structu re-Function Correlations
10.20.3 Substituent Effects 613 with the Leaving Group 651
11 .5.11 Structure-Fun ction Correlations
10.21 The SRN1 Reac tion on Aromatic Rings 615 with the R Group 651
10.21.1 Electron P ushing 615 Effect of the R Group Structure on SN2 Reactions 651
10.21.2 A Few Experimental Observation s 615 Effect of the R Group Structure on SNl Reactions 653
11.5.12 Carbocation Rearran gemen ts 656
10.22 Radical Aro matic Substitutions 615 11.5.13 Anchimeric Assistance in SN1 Reactions 659
10.22.1 Electron Pushing 615 11 .5. 14 SN1 Reactions Involving Non-Classical
10.22.2 Isotope Effects 616 Carboca tions 661
10.22.3 Regiochemistry 616 Norbornyl Cation 662
Cyclopropyl Carbinyl Carbocation 664
Su mmary and Ou tlook 617 11 .5.15 Summa ry of Carboca ti on Stabilization
in Various Reactions 667
EXERC ISES 617 11 .5.16 The Interplay Between Substituti on
FU RTHERREADING 624 and Elimination 667

11.6 Substitution, Radical, Nucleophilic 668
11 .6.1 The SET Reaction-Electron Pushing 668
11 .6.2 The Na ture of the Intermed ia te
in an SET Mechanism 669
CHAPTER 11: O rganic Reaction Mechanisms,
11.6.3 Radical Rea rrangemen ts as Evidence 669
Part 2: Substitutions at Aliphatic 11 .6.4 Structure- Function Correlations
Centers and Thermal Isomerizations/ w ith the Leav ing Gro up 670
Rearrangements 627 11 .6.5 The SRN1 Reaction-Electron Pushing 670
Inten t and Purpose 627
11.7 Radical Aliphatic Substitutions 671
-SUBSTITUTION a TO A CARBONYL CENTER: 11.7.1 Electron Pushing 671
ENOLANDENO LATECH EMISTRY- 627 11.7.2 H ea ts of Reaction 671
11.1 Tautomerization 628 11 .7.3 Regioch emis try of Free Radical
11 .1.1 Electron Push ing for Keto-Enol Halogenation 671
Tautomerizations 628 11.7.4 Autoxidation: Addition of02
11 .1.2 The Th ermod ynamics of Enol Formation 628 into C-H Bonds 673
11 .1.3 Cata lysisofEnoliza tions 629 Electron Push ing for Autoxidation 673
11.1 .4 Kineticvs. Thermodynamic Control - ISOMERIZATIONS AND REARRANGEMENTS- 674
in Enol ate and Enol Forma tion 629
11.8 Migrations to Electrophilic Carbons 674
11.2 a-Halogenation 631 11.8.1 Electron Pushing fo r the
11.2.1 Electron Pushing 631 Pinacol Rearrangemen t 675
11.2.2 A Few Experimenta l Observa ti ons 631 11.8.2 Electron Pushing in the Benzilic Acid
Rearrangement 675
11.3 a-Alkylatio ns 632 11.8.3 Migratory Ap titudes in the Pinacol
11 .3.1 Electron Pushing 632 Rearran gement 675
11 .3.2 Stereochemistry: Conform atio nal Effects 633 11 .8.4 Stereoelectronic and Stereochemical Considerations
in the Pinacol Rearra ngement 676
11.4 The Aldol Reactio n 634 11.8.5 A Few Experimen tal Observations for the Benzilic
11.4.1 Electron Pushing 634 Acid Rearrangement 678
11.4.2 Conformationa l Effects on the Aldol Reaction 634
11.9 Migrations to Electrophilic Heteroatoms 678
-SUBSTITUTIONS O N ALIPHAT IC CENTERS- 637 11 .9.1 Electron Pushing in the Beckmann
11.5 Nucleoph ilic Aliphatic S ubs titution Reactions 637 Rearrangement 678
11 .5.1 SN2 and SN1 Electron-Pushing Exa mp les 637 11.9.2 Electron Pushing for the Hofmann
11.5.2 Kinetics 638 Rearran gement 679
11.5.3 Competition Experimen ts and Product An al yses 639 11 .9.3 Electron Pushing for the Schmidt
11.5.4 Stereochemistry 640 Rearrangement 680
11.5.5 Orbital Considerations 643 11 .9.4 Electron Pushing for the Baeyer-Villiger
11.5.6 Solvent Effects 643 Oxidation 680
11.5.7 Isotope Effect Data 646 11 .9.5 A Few Experimental Observations for the
11.5.8 An Overall Picture of SN2 and SN1 Reactions 646 Beckmann Rearran gement 680


11.9.6 A Few Experimental Observa tions for the Structure-Function Relationships
Schmidt Rearrangement 681 for the Ligands 720
11 .9.7 A Few Experimental Observations for the Oxidative Addition at sp 2 Centers 721
Baeye r-Villiger Oxidation 68J Summary of the Mechanisms for Oxidative
Addition 721
11.10 The Favorskii Rearrangement and Other 12.2.3 Reductive Elimination 724
Carban ion Rearrangements 682 Structure- Function Relationship for the
11.10.1 Electron Pushing 682 R Group and the Ligands 724
11.10.2 Other Carbanion Rearrangements 683 Stereochemistry at the M etal Center 725
Other Mechanisms 725
11.11 Rearrangements Involving Radicals 683 Summary of the Mechanisms for
11.11.1 H ydrogen Shifts 683 Reductive Elimination 726
11 .11.2 Aryl and Vinyl Shifts 684 12.2.4 a- and ()-Eliminations 727
11.J1.3 Rin g-Opening Reactions 685 General Trends for a- and {3-Eliminations 727
Kinetics 728
11.12 Rearrangements and Isomerizations Stereochemistry of{3-Hydride Elimination 729
Involving Biradicals 685 12.2.5 Migratory Insertions 729
11.12.1 Elec tro n Pu shing Involving Biradicals 686 Kinetics 730
11 .12.2 Tetramethylene 687 Studies to Decipher the Mechanism of Migratory
11 .12.3 Trimethylene 689 Insertion In volving CO 730
11.12.4 Trimethylenemethane 693 Other Stereochemical Considerations 732
12.2.6 Electrophilic Addition to Ligand s 733
Summary and Outlook 695 Reaction Types 733
Common Mechanisms Deduced from
Stereochemical Analyses 734
J 2.2.7 Nucleophilic Addition to Ligands 734
Reaction Types 735
Stereochemical and Regiochemical Analyses 735

12.3 Combining the Individual Reactions into Overall
CHAPTER 12: Organotransi tion Metal Reaction
Transformations and Cycles 737
Mechanisms and Catalysis 705
12.3.1 The Nature of Organom etalli c Ca talysis-
Intent and Purpose 705 Ch ange in Mechanism 738
12.3.2 The Mo nsanto Ace ti c Acid Synth esis 738
12.1 The Basics of Organometallic Complexes 705 12.3.3 Hydroformyla tion 739
12.1.1 Electron Counting and Oxidation State 706 12.3.4 The Water-Gas Shift Reaction 740
Electron Counting 706 12.3.5 O lefin Oxidation- The Wacker Process 741
Oxidation State 708 12.3.6 Palladium Coupling Reactions 742
d Electron Count 708 12.3.7 Allylic Alky lation 743
A111bigu ities 708 12.3.8 Olefin Metathesis 744
12.1.2 The 18-Electron Rule 710
12.1.3 Stand ard Geometries 710 Summary and Outlook 747
12.1.4 Terminology 711
12.1.5 Electron Pushing with Organometa llic
Structures 711
12.1.6 d Orbital Splitting Patterns 712
12.1.7 Stabilizing Reactive Ligands 713

12.2 Common Organometallic Reactions 714
12.2.1 Li ga nd Exchange Reactions 714 CHAPTER 13: Organic Polymer and
Reaction Types 714
Materials Chemistry 753
Kinetics 716
Structure-Function Relationships with the Metal 716 Intent and Purpose 753
Struct ure-Function Relationships
with the Ligand 716 13.1 Structural Issues in Materials Chemistry 754
Subs titutions of Other Liga nds 717 13.1.1 Molecular Weight Analysis of Polymers 754
12.2.2 Oxidative Addition 717 Number Average and Weight Average Molecular
Stereochemis try of the Metal Complex 718 Weights-M, and Mw 754
Kinetics 718 13.1.2 Therma l Transi tions-Thermoplastics
Stereochemis try of the R Group 719 and Elastomers 757
Structure-Function Relationship for the R Group 720 13.1.3 Basic Polymer Topologies 759


13.1.4 Polymer-Polymer Phase Behavior 760 SCF Theory 821
13.1.5 Polymer Processing 762 Linear Combination of Atomic Orbitals-
13.1 .6 Novel Topologies-Dendri mers and Molecular Orbitals (LCAO-MO) 821
Hyperbranched Polymers 763 Common Basis Sets- M odeling Atomic Orbitals 822
Dendrimers 763 Extension Beyond HF -Correlation Energy 824
Hyperbranched Polymers 768 Solvation 825
13.1.7 Liquid Crystals 769 General Considerations 825
13.1.8 Fullerenes and Carbon Nanotubes 775 Summary 826
14.2.2 Secular Determinants-A Bridge Between Ab Initio,
13.2 Common Polymerization Mechanisms 779 Semi-Empirical I Approximate, and Perturbational
13.2.1 General Issues 779 Molecular Orbital Th eory Methods 828
13.2.2 Polymerization Kinetics 782 The "Two-Orbital Mixing Problem" 829
Step-Growth Kinetics 782 Writing the Secular Equations and Determinant
Free-Radical Chain Polymerization 783 for Any Molecule 832
Living Polymerizations 785 14.2.3 Semi-Empirical and Approximate Methods 833
Thermodynamics of Polymerizations 787 Neglect of Differential Overlap
13.2.3 Condensa tion Polymerization 788 (NDO) Methods 833
13.2.4 Radical Polymeri za tion 791 i. CNDO, INDO, PNDO (C =Complete,
13.2.5 An ionic Polymerization 793 I= Intermediate, P =Partial) 834
13.2.6 Ca tionic Polymeriza ti on 794 ii. The Semi-Empirical Methods:
13.2.7 Ziegler-Natta and Rela ted Polymerizations 794 M N DO, AMI, and PM3 834
Single-Site Catalysts 796 Extended Hiickel Theory (E HT) 834
13.2.8 Ring-Opening Polymeriza ti on 797 Hucke/ Molecular Orbital Theory (HMOT) 835
13.2.9 Group Transfer Polyme riza tion (GTP) 799 14.2.4 Some General Comments on Computational
Quantum Mechanics 835
Summary and Outlook 800 14.2.5 An Alternative: Density Functional
Theory (DFT) 836
14.3 A Brief Overview of the Implementation
and Results of HMOT 837
14.3.1 Implementing Hucke! Th eory 838
14.3.2 HMOTofCyclic1TSystems 840
14.3.3 HMOT of Linear 1T Sys tems 841
14.3.4 Alternate Hydrocarbons 842
ELECTRONIC STRUCTURE: 14.4 Perturbation Theory-Orbital Mixing Rules 844
THEORY AND APPLICATIONS 14.4.1 Mixing of Degenerate Orbitals-
First-Order Perturbations 845
14.4.2 Mixing of Non-Degenerate Orbitals-
CHAPTER 14: Advanced Concepts in Electronic Second-Order Perturbati ons 845
Structure Theory 807
Intent and Purpose 807 14.5 Some Topics in Organic Chemistry for
Which Molecular Orbital Theory Lends
14.1 Introductory Quantum Mechanics 808 Important Insights 846
14.1.1 TheNa tureofWavefunctions 808 14.5.1 Arenes: Aromaticity and Antiaroma ticity 846
14.1.2 TheSch rodingerEq u ation 809 14.5.2 Cyclopropane and Cyclopropy!carbinyl-
14.1.3 The Hamilton ian 809 Walsh Orbitals 848
14.1.4 The Nature of the ~,72 Operator 811 The Cyclic Three-Orbital Mixing Problem 849
14.1.5 Why do Bonds Form? 812 The MOs of Cyclopropane 850
14.5.3 Planar Methane 853
14.2 Calculational Methods-Solving the Schrodinger 14.5.4 Through-Bond Coupling 854
Equation for Complex Systems 815 14.5.5 Unique Bonding Capabilities of Carbocations-
14.2.1 Ab Initio Molecular Orbital Th eory 815 Non-Classical Ions and Hypervalent Carbon 855
Born-Oppenheimer Approximation 815 Tran sition State Structure Calculations 856
The Orbital Approximation 815 Application of These Methods to Carbocations 857
Spin 816 NMR Effects in Carbocations 857
The Pauli Principle and Determinantal The No rbomyl Cation 858
Wavefunctions 816 14.5.6 Spin Preferences 859
The Hartree-Fock Equation and Two Weakly Interacting Electrons:
the Variational Theorem 818 H2 vs. Atomic C 859 ·


14.6 Organometallic Complexes 862 15.5 Sigmatropic Rearrangements 910
14.6.1 Grou p Orbitals for Metals 863 15.5.1 Theory 911
14.6.2 The Isolobal Analogy 866 15.5.2 Experimental Observations: A Focus on
14.6.3 Using the Group Orbitals to Constr uct Stereochemistry 913
Organometallic Complexes 867 15.5.3 The Mechanism of the
Cope Rea rrangement 916
Summary and Outlook 868 15.5.4 The Claisen Rearran gement 921
Uses in Synthesis 921
EXERC ISES 868 Mechanistic Studies 923
FU RTH ER REA DI NG 875 15.5.5 The Ene Reaction 924

15.6 Cheletropic Reactions 924
15.6.1 Theoretical Analyses 926
CH APTER 15: Thermal Peri cyclic Reactions 877 15.6.2 Carbene Additions 927
Intent and Purpose 877
15.7 In Summary-Applying the Rules 928

15.1 Background 878 Summary and Outlook 928

15.2 A Detailed Analysis of Two Simple EXERC ISES 929
Cycloadditions 878 FURT H ER READING 933
15.2.1 O rbital Symmetry Diagrams 879
[2+2] 879
[4+2 ] 881
15.2.2 State Correlation Diagrams 883
[2+2] 883
[4+2 ] 886 CHAPTER 16: Photochemistry 935
15.2.3 Frontier Molecular Orbital Intent and Purpose 935
(FMO) Theory 888
Con trasting the [2+2] and [4+2] 888 16.1 Photophysical Processes-
15.2.4 Aromatic Transition State The Jablonski Diagram 936
Theory / Topology 889 16.1.1 Electromagnetic Radiation 936
15.2.5 The Generalized Orbital Multiple EnergJJ Surfaces Exist 937
Symmetry Rule 890 16.1.2 Absorption 939
15.2.6 Some Comments on "Forbidden" and 16.1.3 Radiationless Vib ra tional Relaxation 944
"Allowed" Reactions 892 16.1.4 Fluorescence 945
15.2.7 Photochemical Pericyclic Reactions 892 16.1.5 Internal Conversion (IC) 949
15.2.8 Summary of the Various Methods 893 16.1.6 Intersystem Crossing (ISC) 950
16.1.7 Phosphorescence 951
15.3 Cycloadditions 893 16.1.8 Quantum Yield 952
15.3.1 An Allowed Geometry for [2+2) 16.1.9 Summary of Photophysical Processes 952
Cycloadditions 894
15.3.2 Summarizing Cycloadditions 895 16.2 Bimolecular Photophysical Processes 953
15.3.3 General Experimental Observations 895 16.2.1 General Considerations 953
15.3.4 Stereochemistry and Regiochemistry 16.2.2 Quenching, Excimers, and Exciplexes 953
of the Diels-Alder Reaction 896 Quenching 954
An Orbital Approach to Predicting Excimers and Exciplexes 954
Regiochemistry 896 Photoinduced Electron Transfer 955
The Endo Effect 899 16.2.3 Energy Transfer I. Th e Dexter Mechanism-
15.3.5 Experimental Observations fo r Sensitization 956
[2+2) Cycloadditions 901 16.2.4 Energy Transfer II. The Forster Mechanism 958
15.3.6 Experimental Observations fo r 16.2.5 FRET 960
1,3-Dipolar Cycloadditions 901 16.2.6 Energy Pooling 962
15.3.7 Retrocycload ditions 902 16.2.7 An Overview of Bimolecular Photophysical
Processes 962
15.4 Electrocyclic Reactions 903
15.4.1 Terminology 903 16.3 Photochemical Reactions 962
15.4.2 Theoretical Analyses 904 16.3.1 Theoretical Considerations-Funnels 962
15.4.3 Experimental Observations: Diabatic Photoreactions 963
Stereochem istry 906 Other Mechanisms 964
15.4.4 Torquoselectivity 908 16.3.2 Acid-Base Chemistry 965


16.3.3 Olefin Isomerization 965 17.4 Superconductivity 1030
16.3.4 Reversal of Pericyclic Selection Rules 968 17.4.1 Organic Metals /Sy ntheti c Metals 1032
16.3.5 Photocycloaddition Reactions 970
Making Highly Strained Ring Systems 973 17.5 Non-Linear Optics (NLO) 1033
Breaking Aromaticity 974
16.3.6 The Di-1r-Methane Rearrangement 974 17.6 Photoresists 1036
16.3.7 Carbonyls Part I: The Norrish I Reaction 976 17.6.1 Photolithography 1036
16.3.8 Carbonyls Part II: Photoreduction and 17.6.2 Negative Photoresists 1037
the Norrish II Reaction 978 17.6.3 Positi ve Photoresists 1038
16.3.9 Nitrobenzyl Photochemis try: "Caged"
Compounds 980 Summary and Outlook 1041
16.3.10 Elimination of N 2 : Azo Compounds, Diazo
Compounds, Diazirines, and Azides 981
Azoalkanes (1,2- Diazenes) 981
Diazo Compounds and Diazirines 982
Azides 983

16.4 Chemiluminescence 985 APPENDIX 1: Conversion Factors and Other
16.4.1 Potential Energy Surface for a Useful Data 1047
Chemilu minescent Reaction 985
APPENDIX 2: Electrostatic Potential Surfaces for
16.4.2 Typ ical Chemiluminescent Reactions 986
16.4.3 Dioxetane Thermolysis 987 Representative Organic Molecules 1049
APPENDIX 3: Group Orbitals of Common Functional
16.5 Singlet Oxygen 989 Groups: Representative Examples Using
Simple Molecules 1051
Summary and Outlook 993
APPENDIX 4: The Organic Structures of Biology 1057
APPENDIX 5: Pushing Electrons 1061
A5.1 The Rudiments of Pushing Electrons 1061
A5.2 Electron Sources and Sinks for
Two-Electron Flow 1062
CHAPTER 17: Electronic Organic Materials 1001 A5.3 How to Denote Resonance 1064
A5.4 Common Electron-Pushing Errors 1065
Intent and Purpose 1001
Backwards Arrow Pushing 1065
N ot Enough Arrows 1065
17.1 Theory 1001
Losing Tra ck of the Octet Rule 1066
17.1.1 Infinite 1T Systems-An Introduction
Losing Track of Hydrogens and Lone Pairs 1066
to Band Structures 1002
Not Using the Proper Source 1067
17.1.2 The Peierls Distortion 1009
Mixed Media Mistakes 1067
17.1.3 Dop ing 1011
Too Many Arrows-ShortCuts 1067
17.2 Conducting Polymers 1016 A5.5 Complex Reactions-Draw ing a Chemica lly
17.2.1 Conductivity 1016 Reasonable M ech ani sm 1068
17.2.2 Polyacetylene 1017 A5.6 Two Case Studies of Predicting
17.2.3 Polyarenes and Polyarenevinylenes 1018 Reaction Mechanisms 1069
17.2.4 Polyaniline 1021 A5.7 Pushing Electrons for Radical Reactions 1071
Practice Proble ms for Pushing Electrons 1073
17.3 Organic Magnetic Materials 1022 APPENDIX 6: Reaction Mechanism Nomenclature 1075
17.3.1 Magnetism 1023
17.3.2 The M olecul ar Approach to Organic Magnetic Index 1079
Materials 1024
17.3.3 The Pol ymer Approach to Organic Magnetic
Materials-Very High-Spin Organic Molecules 1027


CHAPTER I The Strength of a Buried Salt Bridge 165
How Realistic are Formal Charges? 7 The Angular Dependence of Dipole-Dipole Interactions-
NMR Coupling Constants 10 The "Magic Angle" 168
Scaling Electrostatic Surface Potentials 15 An Unusual Hydrogen Bond Acceptor 169
1-Fluorobutane 16 Evidence for Weak Directionality Considerations 170
Particle in a Box 21 Intramolecular Hydrogen Bonds are Best
Resonance in the Peptide Amide Bond? 23 for Nine-Membered Rings 170
A Brief Look at Symmetry and Symmetry Operations 29 Solvent Scales and Hydrogen Bonds 172
CH 5+-Not Really a Well-Defined Structure 55 The Extent of Resonance can be Correlated with
Pyramidal Inversion: NH 3 vs. PH3 57 Hydrogen Bond Length 174
Stable Carbenes 59 Cooperative Hydrogen Bonding in Saccharides 175
How Much is a Hydrogen Bond in an a -Helix Worth? 176
CHAPTER2 Proton Sponges 179
Entropy Changes During Cyclization Reactions 71 The Relevance of Low-Barrier H ydrogen Bonds
A Consequence of High Bond Strength: to Enzymatic Catalysis 179
The Hydroxyl Radical in Biology 73 13-Peptide Foldamers 180
The Half-Life for Homolysis of Ethane A Cation-'IT Interaction at the Nicotine Receptor 183
at Room Temperature 73 The Polar Nature of Benzene Affects Acidities
The Probability of Finding Atoms at Particular in a Predictable Manner 184
Separations 75 Use of the Arene-Perfluorarene Interaction in the
How do We Know That n = 0 is Most Relevant Design of Solid State Structures 185
for Bond Stretches at T = 298 K? 76 Donor-Acceptor Driven Folding 187
Potential Surfaces for Bond Bending Motions 78 The Hydrophobic Effect and Protein Folding 194
How Big is 3 kcal/ mol? 93 More Foldamers: Folding Driven by
Shouldn't Torsional Motions be Quantized? 94 Solvophobic Effects 195
The Geometry of Radicals 96 Calculating Drug Binding Energies by SPT 201
Differing Magnitudes of Energy Values in
Thermodynamics and Kinetics 100
"Sugar Pucker" in Nucleic Acids 102 CHAPTER4
Alternative Measurements ofSteric Size 104 The Units of Binding Constants 209
The Use of A Values in a Conformational Analysis Cooperativity in Drug Receptor Interactions 215
Study for the Determination of Intramolecular The Hill Equation and Cooperativity in
Hydrogen Bond Strength 105 Protein-Ligand Interactions 219
The NMR Time Scale 106 The Benesi-Hildebrand Plot 221
Ring Fusion-Steroids 108 How are Heat Changes Related to Enthalpy? 223
A Conformational Effect on the Material Properties Using the Helical Structure of Pep tides and the
ofPoly(3-Alkylthiophenes) 116 Complexation Power of Crowns to Create
Cyclopropenyl Cation 117 an Artificial Transmembrane Channel 226
Cyclopropenyl Anion 118 Preorganization and the Salt Bridge 229
Porphyrins 119 A Clear Case of Entropy Driven Electrostatic
Protein Disulfide Linkages 123 Complexation 229
From Strained Molecules to Molecular Rods 126 Salt Bridges Evaluated by Non-Biological Systems 230
Cubane Explosives? 126 Does Hydrogen Bonding Really Play a Role in
Molecular Gears 128 DNA Strand Recognition? 233
Calixarenes-Important Building Blocks for Molecular
CHAPTER3 Recognition and Supramolecular Chemistry 238
The Use of Solvent Scales to Direct Diels-Alder Aromatics at Biological Binding Sites 239
Reactions 149 Combining the Cation-TI Effect and Crown Ethers 240
The Use of Wetting and the Capillary Action A Thermodynamic Cycle to Determine the Strength
Force to Drive the Self-Assembly of of a Polar-TI Interaction 242
Macroscopic Objects 151 Molecular Mechanics / Modeling and Molecular
The Solvent Packing Coefficient and Recognition 243
the 55% Solution 152 Biotin / Avidin: A Molecular Recognition /
Solvation Can Affect Equilibria 155 Self-Assembly Tool from Nature 249
A van't Hoff Analysis of the Formation of a Taming Cyclobutadiene-A Remarkable Use of
Stable Carbene 163 Supramolecular Chemistry 251 XIX


CHAPTERS Pseudo-First Order Kinetics: Revisiting the
Using a pH Indicator to Sense Species Other Cyclopentyne Example 388
Than the Hydronium Ion 264 Zero Order Kinetics 393
Realistic Titrations in Water 265 An Organometallic Example of Using the SSA
An Extremely Acidic Medium is Formed During to Delineate Mechanisms 395
Ph oto-Initiated Cationic Polymerization in Saturation Kinetics Th at We Take for Granted-
Photol ithography 269 SN1 Reactions 397
Super Acids Used to Activate Hydrocarbons 270 Prior Equilibrium in an SN1 Reaction 398
The Intrinsic Acidity Increase of a Carbon Acid Femtochemistry: Direct Characteriza tion of
by Coordination of BF3 276 Transition States, Part I 400
Direct Observation of Cytosine Protonation During "Seeing" Transition States, Part II: The Role of
Triple Helix Formation 287 Computation 401
A Shift of the Acidity of an N-H Bond in Water Due to Th e Use of Pulse Radio lysis to Measure the pK.s
the Proximity of an Ammonium or Metal Cation 288 of Protonated Ketyl Anions 402
Th e Notion of Su perelectrophiles Produced by Discovery of the Marcus Inverted Region 406
Super Acids 289 Using a More O'Ferrall-Jencks Plot in Catalysis 410

Stereoisomerism and Connectivity 300 The Use of Primary Kinetic Isotope Effects to Probe
Total Synthesis of an Antibiotic with a Staggering the Mech anism of Aliphatic Hydroxylation by
Num berofStereocenters 303 lron(III) Porph yrins 425
The Descriptors for the Amino Acids Can Lead An Example of Changes in the Isotope Effect with Varying
to Confusion 307 Reaction Free Energies 428
Chiral Shift Reagents 308 The Use of an Inverse Isotope Effect to Delineate an
C2 Ligands in Asymmetric Synthesis 313 Enzyme Mechanism 431
Enzymatic Reactions, Molecular Imprints, and An Ingenious Method for Measuring Very Small
Enantiotopic Discrim ination 320 Isotope Effects 432
Biological Knots-DNA and Proteins 325 An Example of Tunneling in a Common Synthetic
Polypropylene Structure and the Mass of the U nive rse 331 Organic Reaction 436
Controlling Polyme r Tacticity-The Metallocenes 332 Using Fractionatio n Factors to Characterize Very Strong
CD Used to Distingu ish a-Helices fro m [3-Sheets 335 Hydrogen Bonds 439
Creating Chiral Phosphates for Use as Mecha nistic The Use of a Proton Inventory to Explore the Mechanism
Probes 335 of Ri bonuclease Catalysis 440
A Molecular Helix Created from H ighl y Twisted A Substi tuent Effect Study to Decipher the Reason
Building Blocks 338 for the H igh Stability of Collagen 444
Using a Hammett Plot to Explore the Behavior of a
CHAPTER 7 Catalytic Antibod y 450
Sing le-Molecule Kinetics 360 An Example of a Ch ange in Mechanism in a Solvolysis
Usi ng the Arrhenj us Equation to Determine Differences Reaction Studied Using CJ+ 452
in Activation Parameters for Two Competing A Swain-Lupton Correlation for Tungsten-Bipyridine-
Pa th ways 370 Catalyzed Allylic Alkylation 453
Curvatu re in an Eyring Plot is Used as Evidence for an Using Taft Parameters to Understa nd the Structures
Enzyme Conformational Change in the Cata lysis of Cobaloximes; Vitamin B12 Mimi cs 455
of the Cleavage of the Co-C Bond of Vitam in B12 371 The Use of the Schleyer Method to Determine the Extent of
Where TST May be Insufficient 374 Nucleophilic Assistance in the Solvolysis of Aryl vinyl
The Transition States for SN1 Reactions 377 Tosylates 459
Comparing Reactivity to Selectivity in Free Radical The Use of Swain-Scott Parameters to Determine
H a logenation 378 the Mechanism of Some Acetal Substi tution
Using the Curtin- Hammett Principle to Pred ict the Reactions 462
Stereochemistry of a n Add ition Reaction 379 ATP Hydrolysis-How f3Lc and f3Nuc Values Have Given
Applying the Principle of Microscopic Reversibility Insight into Transition State Structures 465
to Phosphate Ester Chemistry 380 How Can Some Groups be Both Good Nucleophiles and
Kinetic vs. Thermodynamic Enolates 382 Good Leaving Groups? 466
Molecularity vs. Mechanism. Cyclization Reactions and An Example of an Unexpected Product 472
Effective Molarity 384 Designing a Method to Divert the Intermed iate 473
First Order Ki netics: Delineating Between a Unimolecular Trapping a Phosphorane Legitimizes Its Existence 474
and a Bimolecular Reaction of Cyc!opentyne and Ch eckjng for a Comm on Intermed iate in Rhodium-
Dienes 386 Catalyzed Allylic Alkylations 475
The Observa tion of Second Order Kinetics to Support Pyranoside Hydrolysis by Lysozyme 476
a Multistep Displacement Mechanism for a Using Isotopic Scrambling to Distinguish Exocyclic vs.
Vitamin Analog 387 Endocyclic Cleavage Pathways for a Pyranoside 478


Determination of 1,4-Biradical Lifetimes Using C-H Activation, Part II 723
a Radical Clock 480 The Sandmeyer Reaction 726
The Identification of Intermediates from a Catalytic Cycle Olefin Slippage During Nucleophilic Addition to
Needs to be Interpreted with Care 481 Alkenes 737
Pd(O) Coupling Reactions in Organic Synthesis 742
CHAPTER9 Stereocontrol at Every Step in Asymmetric Allylic
The Application of Figure 9.4 to Enzymes 494 Alkylations 745
High Proximity Leads to the Isolation of a Tetrahedral Cyclic Rings Possessing Over 100,000 Carbons! 747
Intermediate 498
The Notion of "Near Attack Conformations" 499 CHAPTER13
Toward an Artificial Acetylcholinesterase 501 Monodisperse Materials Prepared Biosynthetically 756
Metal and Hydrogen Bonding Promoted Hydrolysis An Analysis ofDispersity and Molecular Weight 757
of2',3'-cAMP 502 A Melting Analysis 759
Nucleophilic Catalysis of Electrophilic Reactions 503 Protein Folding Modeled by a Two-State Polymer
Organocatalysis 505 Phase Transition 762
Lysozyme 506 Dendrimers, Fractals, Neurons, and Trees 769
A Model for General-Acid-General-Base Catalysis 514 Lyotropic Liquid Crystals: From Soap Scum to
Anomalous Bmnsted Values 519 Biological Membranes 774
Artificial Enzymes: Cyclodextrins Lead the Way 530 Organic Surfaces: Self-Assembling Mono layers and
Langmuir- Blodgett Films 778
CHAPTERIO Free-Radical Living Polymerizations 787
Cyclic Forms of Saccharides and Concerted Proton Lycra /Spandex 790
Transfers 545 Radical Copolymerization-Not as Random
Squalene to Lanosterol 550 as You Might Think 792
Mechanisms of Asymmetric Epoxidation Reactions 558 PMMA-One Polym er with a Remarkable Range
Nature's Hydride Reducing Agent 566 ofUses 793
The Captodative Effect 573 Living Polymers for Better Running Shoes 795
Stereoelectronics in an Acyl Transfer Model 579 Using 13C NMR Spectroscop y to Eval uate Polymer
The Swern Oxidation 580 Stereochemistry 797
Gas Phase Eliminations 588
Using the Curtin-Hammett Principle 593 CHAPTER14
Aconitase-An Enzyme that Catalyzes Dehydration The Hydrogen Atom 811
and Rehydration 595 Methane-Molecular Orbitals or Discrete Single
Enzymatic Acyl Transfers 1: The Catalytic Triad 604 Bonds with sp3 Hybrids? 827
Enzymatic Acyl Transfers II: Zn(II) Catalysis 605 Koopmans' Theorem- A Connection Between Ab Initio
Enzyme Mimics for Acyl Transfers 606 Calculations and Experiment 828
Peptide Synthesis-Optimizing Acyl Transfer 606 A Matrix Approach to Setting Up the LCAO Method 832
Through-Bond Coupling and Spin Preferences 861
CHAPTER 11 Cyclobutadiene at the Two-Electron Level of Theory 862
Enolate Aggregation 631
Control of Stereochemistry in Enolate Reactions 636 CHAPTER IS
Gas Phase SN2 Reactions-A Stark Difference in Mechanism Symmetry Does Matter 887
from Solution 641 Allowed Organometallic [2 + 2] Cycloadditions 895
A Potential Kinetic Quandary 642 Semi-Empirical vs. Ab Initio Treatments ofPericyclic
Contact Ion Pairs vs. Solvent-Separated Ion Pairs 647 Transition States 900
An Enzymatic SN2 Reaction: Haloalkane Electrocyclization in Cancer Therapeutics 910
Dehydrogenase 649 Fluxional Molecules 913
The Meaning of f3Lc Values 651 A Remarkable Substituent Effect: The Oxy-Cope
Carbocation Rearrangements in Rings 658 Rearrangement 921
Anchimeric Assistance in War 660 A Biological Claisen Rearrangement-The Chorismate
Further Examples of Hypervalent Carbon 666 Mutase Reaction 922
Brorninations Using N-Bromosuccinimide 673 Hydrophobic Effects in Pericyclic Reactions 923
An Enzymatic Analog to the Benzilic Acid Rearrangement: Pericyclic Reactions of Radical Cations 925
Acetohydroxy-Acid Isomeroreductase 677
Femtochemistry and Singlet Biradicals 693 CHAPTER16
Excited State Wavefunctions 937
CHAPTER12 Physical Properties of Excited States 944
Bonding Models 709 The Sensitivity of Fluorescence-Good News and
Electrophilic Aliphatic Substitutions (SE2 and SE1) 715 Bad News 946
C-H Activation, Part I 722 GFP, Part I: Nature's Fluorophore 947


lsosbestic Points-Hallmarks of One-to-One Stoichiometric Photoaffinity Labeling-A Powerful Tool for
Conversions 949 Chemical Biology 984
The "Free Rotor " or " Loose Bolt" Effect on Li ght Sticks 987
Quantum Yields 953 GFP, Part II: Aequ orin 989
Single- Molecule FRET 961 PhotodynamicThera py 991
Trans-Cyclohexene? 967
Retim I and Rh odopsin- The Photochemistry CHAPTER17
of Vision 968 Solitons in Polyacetylen e 1015
Photochromis m 969 Scanning Probe Microsco py 1040
UV Damage o f D NA-A [2 + 2] Photoreaction 971 Soft Lithography 1041
Usin g Photochemi stry to Generate Reactive Interm.e diates:
Strategies Fast and Slow 983


The twentieth century saw the birth of physical organic chemistry-the study of the inter-
relationships between structure and reactivity in organic molecules-and the discipline ma-
tured to a brilliant and vibrant field. Some would argue that the last century also saw the
near death of the field. Undeniably, physical organic chemistry has had some difficult times.
There is a perception by some that chemists thoroughly understand organic reactivity and
that there are no important problems left. This view ignores the fact that while the rigorous
treatment of structure and reactivity in organic structures that is the field's hallmark contin-
ues, physical organic chemistry has expanded to encompass other disciplines.
In our opinion, physical organic chemistry is alive and well in the early twenty-first
century. New life has been breathed into the field because it has embraced newer chemical
disciplines, such as bioorganic, organometallic, materials, and supramolecular chemistries.
Bioorganic chemistry is, to a considerable extent, physical organic chemistry on proteins,
nucleic acids, oligosaccharides, and other biomolecules. Organometallic chemistry traces its
intellectual roots directly to physical organic chemistry, and the tools and conceptual frame-
work of physical organic chemistry continue to permeate the field. Similarly, studies of poly-
mers and other materials challen ge chemists with problems that benefit directly from the
techniques of physical organic chemistry. Finally, advances in supramolecular ch emistry
result from a deeper understanding of the physical organic chemistry of intermolecular in-
teractions. These newer disciplines have given physical organic chemists fertile ground in
which to study the interrelationships of structure and reactivity. Ye t, even while these new
fields have been developing, remarkable advances in our understanding of basic organic
chemical reactivity have continued to appear, exploiting classical physical organic tools and
d eveloping newer experimental and computational techniques. Th ese new techniques h ave
allowed the investigation of reaction mechanisms with amazing time resolution, the direct
characterization of classically elusive molecules such as cyclobutadiene, and highly detailed
and accurate computational evaluation of problems in reactivity. Importantly, the tech-
niques of physical organic chemistry and the intellectual approach to problems embodied
by the discipline remain as relevant as ever to organic chemistry. Therefore, a course in phys-
ical organic chemistry will be essential for students for the foreseeable future.
This book is meant to capture the state of the art of physical organic chemistry in the
early twenty-first century, and, within the best of our ability, to present material that w ill re-
main relevant as the field evolves in the future. For some time it has been true that if a student
opens a physical organic chemistry textbook to a random page, the odds are good that he or
she will see very interesting chemistry, but chemistry that does not represent an area of sig-
nificant current research activity. We seek to rectify that situation with this text. A student
must know the fundamentals, su ch as the essence of structure and bonding in organic mol-
ecules, the nature of the basic reactive intermediates, and organic reaction mechanisms.
However, students should also have an appreciation of the current issues and challenges in
the field, so that when they inspect the modern literature they will have the necessary back-
ground to read and understand current research efforts. Therefore, while treating the funda-
mentals, we have wherever possible chosen examples and highlights from modern research
areas. Fu rther, we have incorporated chapters focused upon several of the modern disci-
plines that benefit from a physical organic approach. From our perspective, a protein, elec-
trically conductive polymer, or organometallic complex should be as relevant to a course in
physical organic chemistry as are small rings, annulenes, or non-classical ions.
We recognize that this is a delicate balancing act. A course in physical organic chemistry XXlll


cannot also be a course in bioorganic or materials chemistry. However, a physical organic
chemistry class should not be a history course, either. We envision this text as appropriate for
many different kinds of courses, depending on which topics the instructor chooses to em-
phasize. In addition, we hope the book will be the first source a researcher approaches when
confronted with a new term or concept in the primary literature, and that the text w ill pro-
vide a valuable introduction to the topic. Ultimately, we hope to have produced a text that
will provide the fundamental principles and techniques of physical organic chemistry,
while also instilling a sense of excitement about the varied research areas impacted by this
brilliant and vibrant field .

Eric V Anslyn
Norman Hackerma n Professor
Uni versity Distinguished Teaching Professo r
University of Texas, Austin

Dennis A. Dougherty
George Grant Hoag Professor of Chemistry
California Institute of Technology

contributing artwork. We especially thank the following for numerous and varied contrib- utions: Bob Bergman. Nick Turro. and providing extensive feed back on some or all of the text. who wrote the accompanying Solutions Manual for the exercises given in each chapter. Nelson Leonard. Kurt Mislow. John Lavigne. Charles Perrin. Acknowledgments Many individuals have contributed to the creation of this textbook in various ways. in- cluding offering moral support. Jane Ellis for stepping up at the right times. John Murdzek for insightful editing. Christine Taylor for or- chestrating the entire process and prodding when appropriate. Timothy Snowden. Suzanne Tobey. Francois Diederich. Anthony Kirby. A very special thanks goes to Michael Sponsler. Grant Willson. Samuel Gellman. Charles Lieber. Akin Davulcu. John Sherman. and Sheryl Wiskur. Tom Webster for dedicated efforts on the artwork. Richard McCullough. Shawn McCleskey. Ken Houk. Wes Borden. XXV . Producing this text has been extraordinarily complicated. Larry Scott. and we thank: Bob Ishi for an inspired design. Finally. Robert Hanes. Jeffrey Moore. We thank Roxanna Anslyn and Ellen Dougherty for their remarkable patience and end- less support. it takes a pair of very understanding wives to put up with a six-year writing pro- cess. and made nu- merous valuable suggestions and contributions. and Bruce Armbruster for enthusiastic support throughout the project. He read each chapter in detail. Scott Silverman has provided numerous corrections and suggestions.


As noted in the Preface. Alternatively. theoretical organic chemis try. There are many possible variations.5-8. and 17 (presumably not in that order) . We also note that many of the highlights make ex- cellent starting points for a five. Further. 11. The order of chapters reflects what we feel is a sensible order of material for a one-year course. The Going Deeper highlights often expand upon an area. and then 14-16 creates a course that is doable in one extremely fast-moving semester. A course that emphasizes structure and theory more than reactivity could involve Chapters 1-6. such as" den- drimer" or "photoresist. mixing in Chapters 3. based upon the interest and expertise of the instructor. we fully recog- nize that at some points the book reads like a " topics" book. along with particularly informative examples we can use as highlights. and 15 for an exhaustive. 14. since we know that physical organic chemistry will continue to evolve and exten d into new disciplines as chemistry tackles new and varied problems. rather than exhaus- tively citing the primary litera ture. In any sequence. Therefore. In our experience. 17. several opportunities for special topics courses or parts of courses are available: computational chemistry. many topics in a more classical physical organic text have been given less coverage herein. although other sequences would also be quite viable. we recognize that at many institutions only one semester. In addi- tion. Either alternative gives a fairly classical approach to the field. or one to two quarters. Finally. In these cases. or to show how the topic being discussed can be put into practice. materials chemistry. 12. 4. We view this book as a "living" text. should enhance the course considerably. dropping Chapters 10 and 11 from this or- der makes a very manageable one-semester course. selected portions of9-11. 13. without a truly in-depth analy- sis of a given subject. We believe we have given enough references at the end of each chapter to enable the instructor to expand any topic. we have chosen to em phasize review articles in the references. one-semester course on thermal chemical reaction mechanisms. but instills the excitement of modern research areas through our use of "highlights" (see below). and perhaps additional chapters as new fields come to benefit from physical organic chemistry. We cannot promise that xxvu . 4. and parts of 6. A Nate to the Instructor Our intent has been to produce a textbook that could be covered in a one-year course in physical organic chemistry. given the breadth of topics addressed. if organic reaction mechanisms are covered in another class. Chapters 2 and 14. Chapters ten-page paper for the student to write. 14-17. We would appreciate instructors sending us suggestions for future topics to cover. The Connections highlights are used to tie the topic at hand to a modern discipline. is devoted to this topic. 12." These are integral parts of the textbook that the students should not skip when reading the chapters (it is probably important to tell the students this). and so on. many instructors may want to consult the primary lit- erature and go into more detail on selected topics of special interest to them. one goal of this text is to serve as a reference when a student or professor is reading the primary literature and comes across unfamiliar terms." However. 13. 10. Chapters 3. covering Ch apters 1-2. but we make a few suggestions here. Given the remarkable literature-searching capabilities now available to most stu- dents. We have d esigned Chapters 9. Chapters 13. and 17 whenever possible. the instructor will need to pick and choose among the chapters and even sections within chapters. or point out what we feel is a particularly interesting sidelight on the topic at hand. supramolecular chemistry. We intend to keep the text current by adding new highlights as appropriate. and each in- structor will likely have a different preferred sequence. One of the ways we bring modern topics to the forefront in this book is through provid- ing two kinds of highlights:" Going Deeper" and" Connections. and perhaps parts of 4.

and even outright errors have crept in. we are sure that some unclear presenta- tions. We wish you the best of luck in using this textbook. misrepresentations. We welcome corrections and comments on these issues from our colleagues around the world. and again welcome suggestions. Many difficult choices had to be made over the six years it took to create this text. but this literature will help us to keep a broad perspective on where the field is moving. . Given the magnitude and scope of this project.xxviii A N OTE TO THE INSTRUCTOR they will all be incorporated. and no doubt the selection of top- ics is biased by our own perceptions and interests. We apologize in advance to any of our col- leagues who feel their work is not properly represented.

Modern Physical Organic Chemistry .




carbonyls. con cepts such as quantum numbers. Showing such a correlation is the third goal of the chapter. and we will often refer to their results during our discu ssions in this and other chapters. we defer detailed discussion s o f sophisticated calcula- tions and full molecular orbital theory until just before the chapters where these methods are 3 . Furthermore. the second goal of the chapter is to present a more ad- vanced view of bonding. We will discuss the method- ology of these calculations in detail in Chapter 14. for all their power. electronega tivity. we must be able to rapidly assess the probability of a reaction occurring without constantly referring to the results of a quantum mechanical calculation. such calcula- tions do not necessarily produce insight into the nature of molecules. p olar cova- lent bonding. molecular conformations and structures. or ki- netics and thermodynamics. We will describe several of these molecular orbitals for many common organic functional groups. should we d evelop simple d escriptive theories of bonding? With the advent of universally available. This more advanced approach to bonding includes the notion of group orbitals for recurring functional groups. A large fraction of organic chemistry can be understood and predicted based upon these very simple concepts in structure and bonding. and even simple alkyl groups such as methylene and methyl-can be viewed as having transferable orbitals. Moreover. In all the discussions there is a single unifying theme. You may be aware that modern computational m ethods can be u sed to describe the bonding in organic molecules. A recurrent them e of this chapter is that organic functional groups-olefins. where computational m ethods are discussed. However. and then apply the lessons to reactive intermediates. We show these bonding models first with stable mole- cules. why not just u se quantum m echanics and computers to describe the bonding of any molecule of interest? In the early twenty-first century. that of d eveloping models of bonding that can be used to explain reactivity. However. it should be clear that our standard models of bonding predict the reactivity and structure of all types of organic structures. A string of computer- generated numbers is just no substitute for a well-developed feeling for the nature of bond- ing in organic molecules. This is known as qualitative molecular orbital theory (QMOT). amides. By covering stable structures alongside reactive intermediates. CHAPTER 1 Introduction to Structure and Models of Bonding Intent and Purpose There are three goals for Chapter 1. and a and 1T bonds. stable and otherwise. practically speaking. we do not need high level calculations and full molecular orbital the- ory to understand most common reactions. and it will lay the foundation for Chapter 14. then. are covered in an introductory m anner. and an extension of molecular orbital theory called perturbational mo- lecular orbital theory that will allow u s to make rational predictions as to how bonding schemes arise from orbital mixing. nearly equivalent from one organic structure to another. In this review. it is true that any desktop computer can perform sophisti- cated calculations on molecules of interest to organic chemists. The first is to review simple notions of chemical bonding and structure. as a preparation for fu- ture chapters. electron configurations. Why. h ybridization. in a typical working scenario at the bench or in a scientific discussion. valence- shell electron-pair repulsion (VSEPR) theory. and stability. very powerful computers. Hence. structure. This review is m eant for readers who h ave a knowledge of atomic and molec- ular structure equivalent to that given in introductory chemistry and organic chemistry text- books.

1. In addition. 1. Most of this material should be quite familiar to you. 3. . .f . We now view the shapes and prope rties of atomic orbi tals as they are obtained from basic quantum mechanics via solution of the Schrodinger equa tion. Also.. and it is typically only the electrons and orbitals of the valence shell that are involved in bonding. nucleic acids. calculations are still severely limited in their ability to address large systems such as proteins. The first is the principal quantum number. Two such models are developed in thi s chapter.2 and 1. The highest electron density is at the atomic nucleus. We use this section to collect the terminology all in one place. The models must be firmly rooted in rigorous theory.1 are a schema tic represen tation of the orbitals in regions of space around the nucleus. Each row in the periodic table indi cates a dif- ferent principal quantum number (with the exception of d and f orbitals. 3. . For n = 1. alternatively described ass. as powerful as they are. 1. w hich takesonpositiveinte- ger values starting with 1 (n = 1. These orbitals are shown in Fi gure 1. The Bohr model of atoms had electrons moving in speci fie orbits (hence the term orbitals) around the nucleus. 2. each row is fu r ther spli t into azimuthal quantum numbers (m = 0.. with decreasing density in all directions in space at increas- s Orbital representations p Orbital representations Figure 1. This limi- tation is even more severe when solvation or solid state issues become critical.1 The general shape of s and p atomic o rbitals for ca rbon.1 A Review of Basic Bonding Concepts In this section we present a number of basic concepts associated w ith chemical bonding and organic structure. For most students. . . 1. . 2. p. which are d ispl aced down one row from their resp ecti ve principal shells). it is still true-and will be true for some time-that descriptive models of bonding that are readily applicable to a wide range of situations are the best way to attack complex problems. . These ca rtoons are the schematics th at chemists typically s ketch. ). This number indicates the angular momentum of the orbital. and mus t stand up to quantitative computational tes ts. and it defines the spatial dis- tribution of the orbital with respect to the nucleus. An atom's highest princi pal quantum number de- termines the valence shell of the atom. The atomic orbital wavefunctions come in sets that are associated with four different quantum numbers.3.. the sh apes and properties of atomic orbitals are of paramount importance in dictating the bonding in and properties of molecules. a quick read of this first section will provide an adequate refresher. The shapes given in Figure 1. or conducting polymers. Shown also is a more realisti c representa tio n for the p orbital produced by qu a ntum mechanica l calculations. Therefore. and to be sure you recall the essentials we will need for the more advanced model of bonding given in Sections 1.1 Quantum Numbers and Atomic Orbitals Every molecule is made up from the nuclei and electrons of two or m ore atoms via bonds that result from the overlap of atomic orbitals. The solutions to the Schrodinger equation are termed wavefunctions. only a 1s atomic orbital is allowed. ). Hence.1 for n = 2 (as with carbon) as a function of one of the three Cartesian coordinates.I ~ 4 CHAPTER 1 I NTRODUCTION TO STRUCTURE AND MODELS OF BO N D I NG essential. and in their most common implementation these wavefunctions correspond to atomic or molecular orbitals. d.

their movem ents are ach1ally correlated. the electron configura tion of carbon is 1s 2 2s 2 2p 2 .lt does not matter which color is defined as p ositive or nega tive.1 A REV l EW OF BASIC BOND I NG CONCEP T S 5 ing dis tances from the nucleus. There is no other m eaning to be given to these phases. and their spin quantum numbers must be oppositely signed (te rmed spin paired) if th e electrons reside in the same orbital. and then drops off.2 Electron Configurations and Electronic Diagrams The electron configuration of an atom d escribes all the atomic orbitals that are popu- lated with electrons.1 (color and gra y / clear) is solely a result of th e mathematical functions describing the orbitals.1.~. 2s. we will examine m etals. the probability of finding an electron in the differently phased regions is the same. Furthermore. The 2s orbital is similar to the 1s or- bital. the most s table form. The aufbau principle (from German for "building-up") states that one populates the lower energy orbitals with electrons first. su ch as 1s 2 2s 1 2p 3 . One color indicates th at the function is positive in this region of space. For exa mple.1. so as to keep the like charges apart. A p orbital ca n orient along three perpendicular directi ons in space. In rea lity. but ha s a spherical surface in three-dimensional space w here the electron density goes to zero. The arrows represent electrons. is associated with the spin of an electron. Correlation is the ability of an electron to feel th e trajectory of another electron and therefore alter its own course so as to minirru ze Coulombic repul sions and keep th e energy of the system to a minimum. Thi s population d ens ity is shown as a dumbbell-like shape. It is specifi ca ll y related to th e square of the m athematical functi on th at represents the orbitals. There is zero probability of finding an electron at an orbital node. The spherical node of a 2s orbital cannot be seen in the representation of Figure 1. The phasing of the atomic orbitals shown in Figure 1. For instance. Promotion of an electron from an atomic orbital to a higher-lyin g atomic orbital produces a higher energy excited state. these waves are overlapping in space when the electrons are in the same orbital. . we pictorially represent this orbital as a sphere. The 2p orbitals ha ve a nodal plane that conta ins the nucl e us and is perpendic- ular to the orbital axis.2. with the number of electrons in each orbita l designated by a super- script. 1. th ey tend to repe l each other. Thi s is th e ground state of carbon. 0.1. Hence. 1. Several rules are used to decide how these lines (orbitals) are populated (fi lled) with elec- trons. and 1. 0.An orbital can only contain two electrons. ca lled a node. only two electrons can be in each or- . Because electrons have wave-like properties. The popula- tion density of a p orbital reaches a ma ximum along its axis in bo th the negative and positive spatial directions. The fourth and final quantum numbe1~ m 5 . In an electronic diagram th e atomic orbitals are represented by horizontal lines at diffe r- ent energy levels. As a result. The principal quantum number 2 has s and p orbitals. carbon h as its 1s. The probabi l- ity is defined as the electron density or electron distribution. and 2. For the p orbita ls the ma gnetic quantum numbers are . and z axes. just as with a 1s orbital. only that the two regions are opposite. A node is a surface (a sphere for s orbitals. and their direction indicates the relative spin of the individual electrons. Its value can be + /1 or. and th e other color indi ca tes that the function is negative. because the electrons are negatively charged and have particle character also. each representing one of the three different orthogona l direction s in space (see th e three 2p orbit- als in Figure 1.1. the e lectron density is zero at the nucle us. However. this nodal surface in the 2s orbital h as little impact on bonding mod els. which contain d orbitals. The directionality of an orbital in space is associated with a third se t of quantum num- bers ca ll ed magnetic. where the higher the line on the page the higher the energy. and 2p orbitals each pop ul ated with two electrons. As such. The 2s and 2p orbitals make up the valence shell for carbon .1). Symbols are placed near the lines to indicate which orbitals the lines are m eant to represent. 1. a plane for p orbitals) that separa tes th e positive and nega tive regions of a wavefunction. defined to be the x. y. We pictoriall y represent such a popul ation density as a sphere. The magnetic quantum numbers ford orbitals are . Later in thi s chapter. and again.

0. In an alternative to the electron dot symbolism. Carbon is said to have a valence of four. even though the charges are in fact delocalized among the atoms in the m olecule. We have to turn to more sophisticated molecular stru cture a nd bond in g concep ts to understand structure and reactivity. In this m e th od. the lowest energy elec- 2s Jl Carbon electron configuration tronic diagra m of the valence shell of carbon is as shown in the margin. d oubl e. such th at each atom achieves an octet of electrons. and all 2p t t electrons in singly-occupied orbitals ha ve their spins aligned. radical. carbanion. or distributions of electrons wi thin molecu les. two. 1. two in the ls orbital. Carbon has six electrons.1. this notation can be used to accurately predict the number of lone pairs th at an atom will have and whether th at atom will use single.1.H H-C= N: I I H H A few examples of Lewis structures The problem with Lewis dot stru ctures is that they provide no insigh t into molecular sh apes. Lewis developed a notation th a t allows us to use the va lence electrons of atoms in a molecul e to predict the bonding in that molecule. and when they are in the same orbital they must be spin paired (a result of the Pauli principle).6 CHAPTER 1 : I NTRODUCTIO TO STRUCTURE A D MODE LS OF BOND I NG bital. we can draw a line to rep resent a bond. the electrons in the va- lence shell of each atom are drawn as dots for all atoms in the molecule (see examples be- low). We singly populate such orbitals sequentially. This valence is by far the most common bonding arrangement for C. N. doubl e. which are orbitals that have the same energy.4 Formal Charge Often it is convenient to associate full charges wi th certain atoms. they are on ly useful for predicting the number of bonds an atom forms. and these full charges on atoms are ca lled formal charges. Instead. carbon w ill make four bonds. orbitals. The num- ber of bonds that an atom can make is called its valence number. w hich s tates that atoms are most s table when their valence she ll is full. When a simil ar analysis is done for N. Despite its simplicity. or carbene. The familiar octet rule. 1. If each bond that carbon makes is crea ted by the donation of a single electron from an adjacent atom's atomic orbitals. w hether the atom has lon e pairs. respectively. Hund's rule tells us how to handle the population of degenerate orbitals. respectively. and one bond(s). and w hether single. no further insight into the structure or reactivity can be obtained from the Lewis structure. denoting they are more of a formality than a reality.H H. namely a carbocation. When carbon has fewer than four bonds it is in a reactive form.H H . Bonds are formed by sharing of one or more pairs of electrons between the atom s. The formula generally give n in introductory chemistry tex tbooks for calculating the for- m al charge is formal charge = number of valence electron s. it is found that these atoms prefer three. and four valence electrons th at occupy the 2s and 2p valence orbitals.3 Lewis Structures G. or triple bonds when inco rpo- rated into specific molecules. H H:C:H H:N:H H:O : H H:C :::N: H H H ·a· I II H-C. or triple bonds are used. while doubl e and tripl e bond s involve the sharing of four and six electrons. Once an atom is found to have an octet us- in g a Lewis analysis. and the overall molecule may be neutra l. A single bond is th e sharing of two electrons. and F.0 -H H.number of unshared electrons . Based on the rules briefly reviewed here. Such charges derive from the Lewis structure.C.N. su ggests that carbon in a molecule will take on four more electrons from other atoms so as to possess an octet of electrons and thereby attain a noble gas configuration.

- ~the number of shared electrons. In organic chemistry, it is easier to just remember a few
simple structures. For example, the oxygen in water has two bonds to hydrogen and is neu-
tral. In contrast, the oxygen in the hydronium ion (H 30 +) bonds to three hydrogens and is
positive; and the oxygen of hydroxide (OH-) has only one bond and has a formal negative
charge. This series can be generalized. Whenever oxygen has an octet of electrons and has
one, two, or three bonds it is negative, neutral, or positive, respectively. More generally,
whenever an atom has an octet and has one bond more than its neutral state it is positive;
when it has one bond fewer it is negative. Hence, a nitrogen atom having two, three, or four
bonds is negative, neutral, or positive, respectively; similarly, a carbon having an octet and
three, four, or five bonds is negative, neutral, or positive, respectively. Although formal
charge can be rapidly evaluated in this manner, you should not take the charge on a particu-
lar atom too literally, as demonstrated in the following Going Deeper highlight.

Going Deeper

How Realistic are Formal Charges? charge on theN. On going from trimethylamine to tetra-
Forma l charge is more or less a bookkeeping tool. For methylammonium theN does become more positive than
the tetramethy lammonium ion, for example, we draw a in a neutral molecule. It is just that it goes from partial neg-
positive charge on the nitrogen because it is tetravalent. ative to esse ntially neutral, rather than from neutral to pos-
However, it is now possible to develop very accurate itive, as implied by the formal charge symbolism. Beyond
descriptions of the electron distributions in molecules bookkeeping, form al charge is really only useful for indi-
using sophisticated computational techniques (Chapter cating the charge on the molecule, not on individual atoms.
14). Such calculations indicate that a much more reason-
CH 3
able model for the tetramethylammoni urn ion describes
theN as essentially neutral. The positive charge resides on H 3 C-~-CH 3 H C - N - CH
3 I 3
the methyls, each carrying one-fourth of a charge. What is
CH 3 CH 3
going on h ere? Looking ahead to Section 1.1.8, we know
that N is more electronegative than C, so it should have Formal charge on
more negative charge (less positive charge) than C. quaternary ammonium
Indeed, in trimethylamine there is a substantial negative

1.1.5 VSEPR
Once we have a basic idea of the bonds to expect for organic structures, the n ex t key is-
sue is the three-dimensional shape of such structures. We now introduce two important con-
cepts for rationalizing the diverse possibilities for shapes of organic molecules: VSEPR and
The valence-shell electron-pair repulsion (VSEPR) rule states that all groups emanat-
ing from an atom-whether single, double, or triple bonds, or lone pairs-will be in spatial
positions that are as far apart from one another as possible. The VSEPR method does not
consider singly occupied orbitals to be groups (see below for the reason). VSEPR is purely a
theory based upon the notion that the electrostatic repulsions between entities consisting of
two or more electrons dictate molecular geometries.
This rule can be applied to carbon when it is bonding to either four, three, or two other
atoms. Acetylene has a linear arrangement of the C-C triple bond and the C-H bond, be-
cause a 180° angle places these two groups as far apart as possible. When three groups are
attached to an atom, such as the three hydrogens of CH3 +, the geometry is trigonal planar

180° 180° H
n n 1 ~109.5 °
..c __ H
=c=o = H-C:=C-H

Geometries based upon VSEPR


with 120° H-C-H bond angles. Finally, in a molecule such as methane, a tetrahedral ar-
rangement of the four bonds places them as far apart in space as possible (H-C-H angles of
The geometries for acetylene, methyl cation, and methane correspond to the bond
angles for idealized linear, trigonal, and tetrahedral systems. While these geometrically per-
fect angles do appear in simple molecules like these, in most organic molecules where differ-
ent groups are attached to the various atoms, measurable deviations from these ideals are
observed. However, we will still loosely refer to the carbons as tetrahedral or trigonal, even
though we don't expect angles of exactly 109S or 120°, respectively.
The VSEPR model provides a simple way to unders tand such deviations from perfec-
tion. Since the geometries derived from VSEPR are based solely upon maximizing the dis-
tance between electron pairs, it makes sense that the geometries would also depend upon
the "sizes" of the electron pairs. A central tenant ofVSEPR is that lone pairs behave as if they
are larger than bonded pairs. Always keep in mind that VSEPR is not based on any first
principles analysis of electronic structure theory. It is a simple way to rationalize observed
trends. It is debatable whether a lone pair of electrons actually is larger than a bonded pair of
·a- electrons plus the associated atoms. In fact, it is not even clear that size is a well defined con-
H/ "H cept for a lone pair. The point is, in VSEPR we consider lone pairs to be larger than bonded
104.SO pairs because that approach leads to the right conclusions. This view allows us to rationalize
the fact that the H-X-H angles in ammonia and water are smaller than 109S. Both systems
Perturbations from a perfect
tetrahed ral angle are considered to have four groups attached to the central atom because, as stated earlier,
lone pairs count as groups in VSEPR. Since a lone pair is larger than a bonding pait~ the N-H
bonds of ammonia want to get away from the lone pair, causing contractions of the H-N-H
angles. The effect is larger in water, with two lone pairs.
The VSEPR rule uses a common principle in organic chemistry to predict geometry, that
of sterics, a notion associated with the through-space repulsion between two groups. Steric
repulsion arises from the buttressing of fill ed orbitals that cannot participate in bonding,
where the negative electrostatic field of the electrons in the orbitals is repulsive. The reason
that singly occupied orbitals are not considered to be groups in VSEPR is that they can par-
Perturbation from the ticipate in bonding with doubly occupied orbitals. Intuitively, we expect larger groups to be
tetrahedral ang le
more repulsive than smaller groups, and this is the reasoning applied to the lone pairs in am-
monia and water. Likewise, due to sterics, we may expect the central carbon in 2-methylpro-
pane to have an angle larger than 109.5°, and indeed the angle is larger than this value (see

1.1.6 Hybridization
It was stated earlier that CH 3 + prefers bond angles of 120°, and methane prefers bond
angles of 109S. How do we achieve such bond angles when the sand p atomic orbitals are
not oriented at these angles? The s orbitals are spherical and so have no directionality in
space, and the p orbitals are oriented at 90° angles with respect to each other. We need a con-
No overlap between
s and p orbital ceptual approach to understand how sand p atomic orbitals can accommodate these experi-
mentally determined molecular bond angles. The most common approach is the idea of hy-
bridiza tion, first introduced by Pauling.
Pauling's assumption was that bonds arise from the overlap of atomic orbitals on adja -
cent atoms, and that the better the overlap the stronger the bond. Orbital overlap has a
quantitative quantum mechanical definition (given in Chapter 14). In a qualitative sense,
overlap can be thought of as the extent to which the orbitals occupy the same space. How-
Increasing ever, if there are regions of overlap with matched and mismatched phasing, the contribu-
overlap tions to the overlap have opposite signs and will cancel. The more space occupied where the
phasing reinforces, the larger the overlap. When the opposite phasing in the various areas
completely cancels, there is no overlap. For example, consider the arrangements of the s and
p orbitals shown in the margin. The top shows how the s and p occupy some of the same
space, but the phasing completely cancels: zero overlap is the result. Any movement of the s
Most overlap orbital to the side increases overlap, until the greatest overlap, shown for the bottom ar-
rangement, takes advantage of the directionality of a p orbital.


0 ±~
s Pr
cf8Px Pz
Da + cQ
2 spHybrids
Px Pz


w 2
3 sp Hybrids

0 ±~
s Py Px Pz 4 sp Hybrids

Figure 1.2
Forming h ybrid orbitals. Combining an s orbital with one, two, o r three p o rbitals
produces the fami liar A. sp, B..sp2, and C. sp 3 h ybrid orbital s.

Pauling also argued that orbitals with directionality would give stronger bonds because
the overlap would be higher. Pauling suggested that to achieve orbitals with directionality,
mixtures of atomic orbitals on the same atom are formed in a process known as h ybridi za-
tion . Hybridization is the method of adding and subtracting atomic orbitals on the same
atom. Remember that orbitals are mathematical solutions to the Schrodinger equation, and
that the addition and subtraction of mathematical equations is ju st an exercise in
is a perfectly valid operation to add orbitals as long as one also does the corresponding
subtraction. Qualitatively, we use the positive and negative phasing along with the three-
dimensional shapes of the orbitals to visualize what the resu It of adding and subtracting the
orbitals would be. For example, Figure 1.2 A shows the result of combining a 2s and one of
the 2p orbi tals, in this case the 2py orbital. Each of the resultant orbitals has a large lobe on one
side of th e atom and a small lobe on the other side, and therefore has greater directionality
than the original orbitals. The addition lead s to an orbital with directionality along the nega-
tive y axis, and the subtraction leads to an orbital with directionality along the positivey axis.
The two lobes ha ve different phasing. The combination of an s orbital with a single p orbital
creates what are called sp hybrid orbitals. Note that these two new orbitals point 180° apart,
as is found in acetylene. Hence, the carbons in acetylene are considered to be sp hybridized
in order to accommodate the experimentally determined geometry. This leaves two pure p
orbitals: Px and Pz·
This addition and subtraction can be carried further to give sp 2 and sp 3 orbitals. Figure
1.2 B shows thi s addition and subtrac tion for sp 2 . Note that th e n ew orbitals are now all120°
apart. The remaining p orbital is perpendicular to the sp 2 hybrid orbitals. It makes good
sense that the mixing of this last p orbital with the sp 2 hybrids would lead to new hybrids that
are above and below the plane formed by the sp 2 hybrid s. In this case four identical orbitals
called sp 3 h ybri d s are the result (Figure 1.2 C). Each points toward the corners of a tetra-
h edron . In organic molecules one of these hybridiza tion states-sp, sp2, or sp3-is invoked
as appropriate when explaining the linear, trigonal planar, or te trahedral geometry of an
atom, respectively.


The geometries for acetylene, methyl cation, and methane correspond to the bond
angles for the different hybridization states sp, sp 2, and sp 3, respectively. Again, most organic
molecules display measurable deviations from th ese ideals, but we still loosely refer to the
atoms as sp, sp 2, or sp 3 hybridized, even though we don't expect angles of exactly 180°, 120°
or 109.5°.
Hybridization provides an alternative "explanation" to VSEPR for such deviations
from ideal angles. In going from pure sp to sp 2, sp 3, and pure p, the angles go from 180° to 120°,
109.5°, and 90°. Thus, decreasing s character leads to decreasing bond angles. We could say
that in ammonja the N-H bonds have lost s character from N relative to a pure sp 3 N, because
the angle is smaller than the perfect tetrahedral angle. In fact, we can quantify this analysis
with a simple relationship. We define a hybridization index, i (Eq. 1.1). Here, the observed
bond angle 8is used in the equation to solve fori.

1 + i cos e= o (Eq. 1.1)

We then define the hybridization as spi. For example, since by definition the tetrahedral
angle is the arc cos(- /3) (- 109S), perfect tetrahedral angles imply i = 3. For ammonia, we
conclude that theN hybrids that bond to Hare sp 34 , and in water the bonds to Hare formed
by sp 4 hybrids. That is, in water the orbitals that make up the 0-H bonds are 80% pin charac-
ter and 20% s, versus the 75:25 mixture implied by sp 3 . The lone pairs must compensate, and
they take on extras character in H 3 and H 20. We will see that this notion of non-integra l hy-
bridizations is more than just an after-the-fact rationalization, and has experimental support
(see the following Connections highlight).It can have predictive power. However, we must
first introduce another importa nt bonding concept: e lectronega ti vi ty.


NMR Coupling Constants This correspondence is ind eed seen from an an alysis of
The view of variable hybridi za tion has some experimental the C-H coupling constants g iven be low. The smaller
support. The magnitud e of 13 C- 1H NMR coupling con- rings have the larger coupling constants.
stants is expected to be pro portional to the amount of car-
Ring system /t JC-11-J
bon 5 character in the bond, because only 5 orbitals have
density at the carbon nucleu s a nd can affect neighboring Cyclopropane 161
nuclear spin states. in severa l systems, a clear correlation Cyclobu tane 134
has been observed betwee n NMR coupling constants and Cyclopentane 128
percent 5 character, as predicted from the geometry and
Cyclohexane 124
the associated hybridi zation ind ex.
Cycloheptane 123
For example, in cycl ic a lkanes, the smaller the ring,
the larger the p character that would be expected in the Cyclooctane 122
hybrid orbi ta ls used to form the C-C bonds, because p
orb ita ls better accommodate smaller bond angles. Corre- Fe rg uson, L. N. {1973). Highlights of Alicyclic Chemistry, Part'/ , Franklin
spondingly, the C-H bond s wou ld ha ve higher 5 character. Publishing Company, ln c., Pa li sa de, MI.

1.1.7 A Hybrid Valence Bond/Molecular Orbital Model of Bonding
There are two dominant models for considering bonding in organic molecules: valence
bond theory (VBT) and molecular orbital theory (MOT). While often viewed as competing
theories, VBT and MOT actually complement each other well, and our ultimate model for
bonding will borrow from both theories. VBT was developed first. The idea, as orig inally
put forth by Heitler and London and expanded by Pauling, was that th e binding energy be-
tween two atoms arises primarily from exchange (resonance) of electrons between the two
atoms in a bond. The starting point for VBT has one electron on each atom that contributes
to an electron pair is this assignment of electrons primarily to individual atom s-ot~
more precisely, to individual orbitals on atoms-that is the hallmark of VBT. Bonding is, in


effect, viewed as a perturbation of this arrangement. That is, when two atoms are brought to-
gether, each electron is permitted to interact with either nucleus, and this produces bond en-
ergies in adequate agreement with the experimental values. Hence, the conclusion was that
bonds consist of two electrons in the region between two nuclei.
In VBT a molecule is formed by adjacent atoms sharing electrons. As suggested by the
name, the electrons that are involved in bonding are those from the atoms' valence shells.
Each atom donates one electron to the bond, and the resulting electron pair is considered to
be mostly localized between the two adjacent atoms. This localization of the electrons is ex-
actly the impression of bonding that is given by a Lewis structure. Furthermore, localization
of the electrons between the atoms would require orbitals that point in the appropriate direc-
tions in space. It is this kind of reasoning that led Pauling to develop hybrid orbitals, an es-
sentially valence bond concept. In essence, VBT nicely encompasses the topics discussed to
this point in the chapter. However, one other notion is required by VBT-that of resonance.
As discussed in Section 1.1.10, if more than one Lewis dot structure can be drawn for a mole-
cule, then VBT states that the actual molecule is a hybrid of these" canonical forms".

Creating Localized u and 1r Bonds
The most common model for bonding in organic compounds derives from VBT and the
hybridization procedure given previously. Sigma bonds (CJ bonds) are created by the over-
lap of a hybrid orbital on one atom with a hybrid orbital on another atom or an s orbital on
hydrogen (Figures 1.3 A and B, respectively). Pi bonds ('IT bonds) are created by the overlap
of two p orbitals on adjacent atoms (Figure 1.3 C). Specifically, CJ bonds are defined as having
their electron density along the bond axis, while 'IT bonds have their electron density above
and below the bond axis. The combination of the two orbitals on adjacent atoms that creates
in-phase interactions (signs of the orbitals are the same) between the two atoms is called the
bonding orbital. The combination that results in out-of-phase interactions (signs of the or-
bitals are opposite) is called the anti bonding orbital. The bonding orbital is lower in energy
Nodal plane
than the antibonding orbital. There are also orbitals that contain lone pairs of electrons,
which are not bonding or antibonding. These are called nonbonding orbitals. In standard
neutral organic structures, only the bonding orbitals and nonbonding orbitals are occupied ,- , \lJn*
with electrons. Recall that an alkene functional group has a single u and a single 'IT bond be-
tween the adjacent carbons, whereas an alkyne has a single CJ and two 'IT bonds between the
: ,_00
carbons. The number of bonds between two atoms is called the bond order.
The creation of bonding and antibonding orbitals is actually a molecular orbital theory
notion. Therefore, the orbitals of Figure 1.3 are in effect molecular orbitals. We will have
much more to say in Section 1.2 about how to linearly mix orbitals to create bonding and
antibonding molecular orbitals. However, you may recall molecular orbital mixing dia-
grams from introductory organic chemistry, such as that shown in the margin for the 'IT bond
in ethylen e. These diagrams give a picture of how chemists visually create bonding and anti-
bonding orbitals via mixing. The mixing to derive the molecular orbitals gives both a plus Orbital mixing diagram

A. ~ ~
Bonding Antibonding
orbital orbital
Figure 1.3
B. ~ ~ A. Combination of two hybrid orbitals on adjacent atoms gives bonding
Bonding Anti bonding
and anti bonding orbitals. Population of the bonding orbital with two
orbital orbital electrons creates au bond. B. Combination of a hybrid and a ls orbital
on hydrogen gives a bonding and anti bonding pair. Population of the

~ ~
bonding orbital with two electrons creates au bond. C. Combination of
c. two p orbitals on adjacent atoms also gives a bonding and antibonding
set. Population of the bonding orbital with two electrons creates a TI bond.

Bonding Anti bonding
orbital orbital


and minus combination of the starting atomic orbitals. Note, therefore, that the orbitals
given in Figure 1.3 are actually derived from a h ybrid VBT /MOT approach to bonding. One
creates discrete, localized bonds between adjacent atoms as pictured with VBT, but it is done
u sing the linear combination ideas of MOT.
The simple molecular orbital mixing diagram given above serves to illustrate many con-
cepts and terms used with molecular orbital theory. The bonding molecular orbital (MO) is
symmetric with respect to a mirror plane that resides between the two carbons making the
bond, while the antibonding MO is antisymmetric. Furthermore, a nodal plane exists in the
antibonding MO between th e two atoms making the 'lT bond, which means that populating
this orbital with electrons leads to a repulsive interaction between the atoms.
The picture of cr and 'lT bonds that consist of bonding and antibonding molecular orbitals
that reside primarily be tween adjacent atoms is the standard that organic chemists most
commonly use. The reactivity of the vast majority of organic compounds can be nicely mod-
eled using this picture, and it forms the starting point for the electron pushing method of
presenting organic reaction mechanisms (see Appendix 5). Hence, this theory of bonding is
extremely important in organic chemistry.

1.1.8 Polar Covalent Bonding
Once the geometry of a molecule has been established, the next crucial feature for pre-
dicting the reac tivi ty is its charge distribution. Notions such as VSEPR and hybridization
control shape and structure. Here we discuss h ow electron ega tivity is the primary determi-
n ant of the charge distribution in a molecule, with hybridization playing a secondary but
still imp ortant role.
Covalent bonds predominate in organic chemistry. In our simple theory of bonding, the
two electrons in the bond are sh ared between the two adjacent atoms, as implied by a Lewis
dot structure and the cr and 'lT bonds discussed previously. Very few, if any, organic structures
can be considered to have ionic bonds. However, whenever a carbon forms a bond to any
atom or group n ot identical to itself, the bond develops some polar character; there is a pos-
itive end and a negative end to the bond. This ch arge separa tion m eans the sharing of
electrons is unequal. A covalent bond that h as an unequal sharing of the bonding pair of
electrons is called polar covalent. Pauling argued that introducing polarity into a bond
strengthens it, and we will see in Chapter 2 th at trends in bond s trengths generally su pport
this view.

To predict the charge distribution in an organic molecule, we n eed to examine the elec-
tronegativity of the a toms in the molecule. Pauling original! y d eveloped this important con-
cept and described it as "the power of an atom in a molecule to attract electrons to itself".
Pauling assigned va lues to various atom types by examining bond dissociation en ergies of
molecules. As such, the Pauling electronegativity scale depends upon molecular properties,
and is not an intrinsic property of the atoms. Th e Pauling scale is most commonly u sed, and
is given in all introductory chemistry textbooks.
Mulliken defined an electronegativity scale that is derived from the average of the ion-
ization potential and e lectron affinity of an a tom, and therefore is solely an atomic property.
Th e ionization potential is the energy required to remove an electron from an atom or mole-
cule. Hence, this number reflects the affinity of an atom for th e electrons it already has. The
electron affinity is the amount of energy released or required to attach another electron to
an atom or molecule. Hence, this number reflects the affinity of the atom for an additional
electron. Using these values is a logical basis for determining the ability of an atom to at-
tract electrons toward itself. Along with the electronega tivity scales of Pauling and Mulli-
ken, comparable scales h ave been deve loped by Nag le, Allen, Sanderson, Allred-Rochow,
Gordy, Yuan, and Parr. Suffice it to say that the electronegativity of an atom is a difficult con-
cept to put a precise number on, and that the use of different scales is appropriate for dif-
ferent applications. Table 1.1 compares the Pauling and Mulliken electronegativity scales,
showing that the two are similar. We should always remember that the key issue is the rela-

Table 1.1
Electronegativities of Atoms According to
the Scales of Pauling and Mulliken*
Atom Pauling Mulliken
H 2.1 3.01
B 2.0 1.83
c 2.5 2.67
N 3.0 3.08
0 3.5 3.22
F 4.0 4.44
CI 3.0 3.54
Br 2.8 3.24
I 2.5 2.88
Li 1.0 1.28
Na 0.9 1.21
K 0.8 1.03
Mg 1.2 1.63
Ca 1.0 1.30
AI 1.5 1.37
Si 1.8 2.03
p 2.1 2.39
s 2.5 2.65

' Pauling, L. (1960). Tlte Natureoftltc Cltclllicnl Baud and lite
Structu re of Molewles nud Cn;stals; au lutroductio11 to Modem
Structural Clte111istry, 3d ed., Cornell Uni versity Press, Ithaca,
NY, and A lle n, L. C. "Electronegativity is the Average O ne-
Electron Energy of the Va lence-She ll Electrons in Ground-
State Free Atoms." f. A 111. CilL'/11. Soc., 111, 9003 (1989).

tive electronegativities of two moie ties- we mainly want to know which is the more electro-
negative of the two, and whether the difference is relatively large or small. In mos t situa-
tions, all the various electronegativity scales lead to the same predictions.
The major factor infl uencing electronega ti vity is the energy of the orbitals th a t the atom
uses to accept electrons. As one moves left-to-right across the p eriodic table, the va len ce or-
bitals become lower in energy within the sam e row. Going down a column, th e atoms get
bigger and the valence orbitals are higher in energy. This would m ean that He has the lowest
energy valence orbitals, but He cannot accept an y m ore electrons because it is a noble gas.
The atom with the lowest energy valence orbitals that is not a noble gas is F, which is the most
electronegative e lement. In fact, a useful way to estimate electronega tivities if Table 1.1 is u n-
ava il able is simply to recall that F is the most electronegative elem ent, and moving left or
down in the periodic table progressively diminishes electronega tivity.
When an a tom w ith a higher electronegativity than carbon form s a bond to carbon, the
e lectrons in th e bond will reside more toward the electronega ti ve a tom, producing a partial
negative charge on this atom and a partial positive charge on carbon. Conversely, w h en the
atom has a lower electronegativity, the carbon will possess a p artial negative ch arge. These
partial charges are denoted .5+ and .5- (see a few exa mples below). The larger the difference
between the electronegativities for atoms in a bond, the more the bond is polarized. The
magnitude of the polarization can be gauged using the values of Table 1.1. For example, us-

80 080

H1 '"j
' H H3 C
_..-- C ,
88 CH3
Partial charges



ing the Pauling scale, a C-F bond is more polar than a C-Cl bond, because the electronega-
tivity differences for the atoms in these bonds are 1.43 and 0.61, respecti vely. A bond w hose
electronega ti vity difference is 1.7 is considered to be 50% ionic and 50% covalent, and so
bonds with differences grea ter than this are considered to be ionic. In this view, bond s of Li,
Na, or K to F, Cl, Br, or I are all ionic.
Some conclu sions that ca n be drawn from the data of Table 1.1 m ay seem a bit counterin-
tuitive . For example, the electronega tivity difference between C and I in both sca les is actu-
ally smaller th an the difference between C and H . Therefore, a C-I bond is predicted to have
a smaller charge polarization than a C-H bond, based solely upon electronega ti vities. This
polariza ti on may com e as a surpri se, sin ce iodide is a good leaving group in SN2 reactions.
Often the electron ega tivity of iodine is erroneously in voked to explain such reaction s. As we
will see below, electronega tivity is no t the whole story. Polarizability is also important, esp e-
cially w hen ra tionalizing reactivity trend s. Also note that the electronegativity difference be-
tween C and 0 is smaller than between P and S, or P and 0, on the Mulliken scale. Hence,
S-0 and P-0 bonds are more polarized than C-0 bond s.

Electrostatic Potential Surfaces
In a polar bond, one end is designated as 5+ and the other as 5-, and this design ation is
often adequate for discu ssing simple molecules such as methyl chloride. However, in more
complex mol ecules there will be many different types of bond s with differing degrees of po-
larity, and the overall molecule w ill reflect the sum of these and any interactions they experi-
ence. The simple 5+ I 5- symbolism is no longer adequate, so we need an alternative way to
view the charge distribution in complex organic molecul es. In recent years, m an y scientists
have found that plots of electrosta tic potential surfaces are quite useful in this regard . Such
plots are given in Appendix 2, which contains a gallery of representa tive electrostatic poten-
tial surfaces for prototype organic structures. In these pictures, red represents negative elec-
tros tati c potential whereas blue represents positive electrostatic potential. A green color is a
region that is essentially neutral.
Let's examine for a moment the electrostatic potential surface for meth yl aceta te (see
Appen d ix 2). In our 5+ / 5- method , we would denote methyl acetate as shown in the ma r-
gin. The carbon yl carbon is partially positive, and so is the methyl group. However, bo th ox-
ygens would be d enoted as nega tive. This notation is less than optimum, however, beca use
Partial charges
we h ave no idea w hich oxygen is more nega ti ve. Yet, a quick glance at the electrostatic sur-
face shows that the carbonyl oxygen is more nega ti ve (you may p redict this fac t from reso-
nance; see Section 1.1.10). Thu s, an e lectrosta ti c potential surface can provide deeper insight
into the electronic distribution in a mo lecule than a simple 5+ I 5- picture.
What exactly are these surfaces? First, they result from a full quantum mechanical calcu-
la tion of the electronic structure of the molecule. Note that we do not use electronegativities
or h ybridiza tions or bond dipoles or an y of the d escripti ve features of our bonding mod el in
such ca lcul ations. These are a priori quantum mechanica l calcul ati.ons, and their output en-
ables the charge distribution in a m olecule to be computed.
What are we actually showing in su ch plots? Firs t, we give a surface to the molecu le. The
surface is very similar to a van der Waals surface, the surface that would be obtained by con-
siderin g each atom to be a sphere with a radius equal to its van der Waal radius (Table 1.5).
For technical reasons, though, the surfa ce is more typically an isodensity surface, m eanin g a
surface with a cons ta nt electron density, such as 0.002 electrons / A2 • The di stin ction is sma ll.
Next, we color the surfa ce according to electrostatic potential (i.e., red for nega tive and blue
for positive).
What is electrosta ti c potenti al and how is it d e termined ? Imagine taking a very sm all
sphere with a charge of + 1 and rolling it around the isodensity surface. A t each point, we ask
w hether the sphere is attracted to or repulsed by the surface and w hat the energetic magni-
tude of the interaction is. The magnitude of the interaction is the electrostatic potential.
Thus, the plots have units not of charge but of energy (we will use kcal / mol) . We are not
plo tting partial charge; we are plotting electrostatic p o tenti al, although the two will track

each o ther. As a result, these plots are extremely useful for visualizing the charge distribu-
tion in organic molecules. We encourage the student to consult Appendix 2 frequently while
reading this text.
There are definitely some caveats to the interpretation of electrostatic surface potentials,
and so m e are given in the next Going Deeper highlight. An important caveat is to appreciate
tha t th ese are electros tatic potential surfaces for the ground states of the molecules. They
sh ow the charge distribution absent any external perturbation. When a chemical reaction oc-
cu rs, we expect a substantial reorganization of charge. For example, when an anionic nu-
cleophil e adds to the carbonyl of acetone, we expect a very different electrostatic potential
surface for the transition state than in the picture of acetone shown in Appendix 2. Since
transition states control reactivity, it is risky to use these electrostatic potential surfaces to
predict or ra tionalize reactivity. They can be helpful in this regard, but caution is in order.

Going Deeper

Scaling Electrostatic Surface Potentials kcal l mol. A potential problem with this approach is that
Electrostati c potenti al surfaces are very useful guides to the color scale is "linea r" from plus to minus, so the zero
charge distributi ons, and they are now commonly shown electrostatic potential color-an important benchmark-
in introd uctory organic texts. However, some caution is will be different for this structure than for some other
warranted when interpreting them . Most important is to structure with a range such as + 27.22 to -49.83. For this
pay attenti on to the energy scale associated with any par- reason, we avoid such presentations in Appendix 2 and
ticular structure. That is, what value of positive electro- present electrostatic potential surfaces with a symmetrical
stati c potenti al does it take to achieve the maximum in range of electrostatic potentials, such as ::':: 25 kcal l mol or
blu e, and what negati ve electrostatic potential will maxi- ±50 kcal l mol. In this way, zero electrostatic potential is
mize the red? Rea li ze that in the analysis of any electro- always the same green color. However, it is also very impor-
stati c potenti al surface, positive or negative potentials tan t to be aware of the range of electrostatic potentials being plot-
larger in magnitude th an the arbitrarily set range will ted. A plot of the benzene electrostatic potential surface
simpl y be the most intense blue or red, and will not be with a ::':: 25 kcal I mol range will look different from a plot
di stingui shed from any other value over the limit. with a ::':: 50 kcal I mol range. It is especially risky to com-
There are two ways to scale these plots. Some elec- pare two structures with different ranges plotted. When-
trostatic potential surface presentations take the direct ever possible, we will provide comparisons with the same
results of the calcul ation and use as the maxima the maxi- range, and we will always make it clear what range of elec-
mum valu es for plus and minus electrostatic potential trostatic potentials is used for any figure. The student
in the mol ecule. Thi s appears to be the most common always need s to be aware of electrostatic potential range
approach in introductory texts.ln this approach, we could when interpreting and comparing such plots.
end u p with a range such as + 57.29 kcal l mol to -36.43

Inductive Effects
We h ave seen that when carbon bonds to an electronegative element like 0, N, Cl, or F, a
bond polariza tion develops, making the C 8+ and the heteroatom or halogen 8-. We might
exp ect a functional group containing electronegative atoms to also be electron withdrawing
(see examples in the margin). The phenomenon of withdrawing electrons through a bonds
to the more el ectronega tive atom or group is called an inductive effect. It is an effect that we Partial charges due to induction

will often cite to explain trends in thermodynamics or reactivity throughout this book (an in-
teresting twist to induction is given in the next Going Deeper highlight). The inductive effect
is what gives rise to bond polarizations, polarizations within molecules, and bond and mo-
lecular dipole moments .
A similar but separate phenomenon is a field effect. This is a polarization in a molecule
that results from charges that interact through space, rather than through a bonds, and it can
influence the structure and reactivity of other parts of the molecule. We will see systems later
in which both field and inductive effects seem to be operative.


Going Deeper

1-Fluorobutane negative charge, o n the order of - 0.03. Rather than a p ro-
Consider the specific case of 1-fluorobutane. We expect a gressively dim ini shing positive charge as we mo ve down
large bond dipole, with C1 p ositi ve, and the F nega ti ve. the chain, a charge alterna tion is seen . At some levels of
Moving al ong the u ch ain w e would expect the magnitu de theory, this charge alternation continues down the chain,
of the charges to p rogressively diminish, with som e polar- but the magnitude of the charges a t C3 and C4 are so sm all
ization at C2, and less a t C3, etc. as to be inconsequential.
Inductive effe cts have been extensively investiga ted Thi s result is fairly general in computation al studies,
expe rimentall y, and some conflicting trends are seen . but it is not in line with ex perimental observatio ns of
Ad va nced quantum m ech anical calcul ations of the sort ind uctive effec ts. As we will see in Chapter 5 and other
described in Chapter 14 are able to assign par ti al ch arges pl aces, ind uctive effects on thermod ynamics and kinetics
to atoms or groups of atom s, and while there is some do not usu ally show an altern a tion pa ttern . For example,
deba te as to the best w ay to do this, all m ethod s produ ce for a linear alka noic acid, adding a strongl y electronega-
simil ar trend s. In the case of 1-fluorobutane the F a tom car- tive element like F to the alkyl ch ain always increases the
ries a large parti al negative ch arge, on the order of-0.44. acidity of the carboxylic acid function al gro up, and the
As expected, the CH 2 group at C1, as a unit, carries a com- effect is always stronger the closer the F is to the incipient
parable positive charge, creating a very large C-F bo nd carboxyla te. This trend is an example of the danger of
dipole. What aboutth e CH 2 a t C2? Instead of being jus t d irectly ex trapolating ground s ta te electronk structure
part of a simple C-C bond, it is paired to a carbo n w ith a fea tures to reac ti vity p a ttern s.
substantial positive ch arge. At the same time, it is two Mag nitude of the delta charges
atoms away from a very electronegative ele men t- dimini shes but alternates
fluorine. 80 oG
Perhap s surprisingly, the C2 CH 2 shows a sm all ~
oG F

Group Electronegativ ities
It is often convenient to consider groups tha t m ake up p articular portions of a molecule
as h av ing their own electron egativity. For example, we would expect a CF3 group to affect
the charge distribution in a m olecule via induction mu ch more than a CH 3 group, but if w e
consid er onl y carbon electron ega tivities, the two are the same. Table 1.2 li sts some group
electron egativity valu es that were derived to be comparable to the Pauling scale for atoms.
We find that a methyl group is essen ti ally the sam e as a C, whereas the CF3 group has an elec-
tronegativity similar to th a t for 0. Alkenyl and alkynyl groups are quite electronegative, as
are nitro and cyan o grou ps. Finally, a full p ositive ch arge, su ch as that associated with a pro-
tona ted am ine, h as the hi gh es t grou p electron ega tivity.

Gro up Electronegativities, Scaled to be
Compatible with the Pauling Scale*
G roup Electron egativity
CH3 2.3
CH2CI 2.8
CHCh 3.0
CCb 3.0
CF 3 3.4
Ph 3.0
CH = CH 2 3.0
C = CH 3.3
C= N 3.3
NH 2 3.4
NH 3 + 3.8
N0 2 3.4
OH 3.7
*Wells, P.R. "Grou p Electronegativities." Prog. Phys .
Org. Chem., 6,111 (1968).


Hybridization Effects
The relative electronegativities of C and H, a critical issue in organic chemistry, has in
fact been the topic of some d ebate. In Table 1.1 we see that the Pauling scale describes Cas
more electronegative than H, while the Mulliken scale gives the opposite ordering. It is now
believed that the cause of this discrepancy is a hybridiza tion effect. Since s orbitals have sub-
stantial density at the nucleus while p orbitals h ave a node at the nucleus, the more s charac-
ter in a hybrid orbital, the closer to the nucleus the electrons in that hybrid tend to be. Be-
cause electronegativity describes an atom 's ability to attract electrons to itself, sp 2 hybrids
should be more electronegative than sp3 hybrids, and this is indeed the case. The d a ta in
Table 1.2 are completely consistent with this view. The electronegativities are C=CH >
CH = CH 2 > CH 3 ;sp > sp 2 > sp 3 •
What does this mean, then, about the relative electronegativities of C and H? A good
deal of evidence, including molecular quadrupole moments discussed below, points to the
conclusion that an sp 2 Cis more electronegative than H, while an sp3 C and H have very simi-
lar electronegativities. In a sense, both Pauling and Mulliken were right. A great many ob-
servations, especially those involving noncovalent interactions (Chapters 3 and 4), can be
understood from this simple statement.
Taken together, electronegativity and h ybridi zation provide an appealing rationaliza-
tion of many structural trends . For example, the smaller bond angles in ammonia and water
vs. methane discussed previously are nicely explained. There is a competition as to whether
the central atom (0 or N) should place mores character in the hybrid orbital that contains the
lone pair(s) or the hybrid orbital used to make bonds to the hydrogen atoms. Since the lone
pair electrons are not shared with another atom, an electronegative element prefers greater s
character in its own lone pair orbitals, because it can better keep these electrons to itself. This
effect places more p orbital character in the bonds to hydrogen, which in turn reduces the
H-N-H or H-0-H bond angles relative to methane. The effect is more pronounced for 0 be-
cause it is more electronegative.
As another example, let's consider methyl fluoride. The H-C-F angle is contracted, and
as a result the H-C-Hbonds are slightly expanded. The H-C-F contraction is due to the fact 1.38 A
that F is the more electronega tive subs titu ent. The F prefers to bond to a carbon hybrid that 1 ogA
yI '\ 108.2°
has more p character, because it is easier to withdraw electrons from a p orbital on carbon '\. .C .__
H''~ H
than an s orbital on carbon. It is often said that s orbitals have better electron penetration to H
the nucleus than p orbitals, suggesting again that it is harder to withdraw electrons from s or- 110 .2°
bitals. If the carbon uses more p character in a hybrid to bond to F, more s character will be de- Geometry of CH 3 F
voted to the hybrids that comprise the H-C-H bonds. It is difficult to imagine a rationaliza-
tion of this result using VSEPR, because F is larger than H, and may be expected by VSEPR to
open up the H-C-F angle. For the most part, organic structures are be tter rationalized using
hybridiza tion and electronegativity arguments than VSEPR.

1.1.9 Bond Dipoles, Molecular Dipoles, and Quadrupoles
One goal of our discussion of electronegativity was to delineate the relative electron
withdrawing nature of an atom, group, or orbital. The term "relative" is important, because
the exact numbers associated wi th atom and group electronegativities are not used on a day
to day basis when practicing organic chemistry. Instead, the trends and relative electron do-
nating and accepting abilities are of paramount importance. Now, however, we will be con-
sidering bond dipoles, and we need to get more quantitative. The exact charges on the atoms
in the bonds need to be known, or the electronegativity numbers associated with atom s need
to be used.

Bond Dipoles
Wh en two atoms of differing electronegativites are bonded, one end of the bond will be
5+ and the other will be 5-. This analysis leads to the notion of a bond dipole as the local
moment that is associated with a polar covalent bond. A moment reflects the electrostatic
force that would be exerted by a charge on a neighboring charge. The dipole moment pro-


vides a means of comparing which bonds are more polar and evaluating the relative force
that a dipole exerts on neighboring charges or dipoles. Certainly, examination of the electro-
static potential surfaces of Appendix 2 reveals bond dipoles. If we know the partial charges
on the atoms of the bond, we can calculate a bond dipole.
A dipole moment (J.i) is given in units of electrical charge times distance (Eq. 1.2), where
qis charge and r is distance. It is usually expressed in units of Debye (0, where 10 = IQ- 18 esu
em). "Esu" stands for" electrostatic unit", and the charge of an electron or proton is negative
or positive 4.80 X IQ- 10 esu , respectively.

J.1 = q x r (Eq. 1.2)

For example, a bond that has a 0.2 positive and negative charge on the opposite ends with a
separation of 1.54 A (1 A = 10- 8 em) would have an associated bond dipole of (0.2)(4.80 X
Bond dipole moments IQ- 10 esu)(l.54 x IQ- 8 em) = 1.47 X IQ- 18 esu em = 1.47 D. Chemists consistently use a sym-
bolism in which the positive end of the dipole is represented by a cross, along with an arrow
that points in the direction of the negative end of the dipole.
Recognizing polar covalent bonds and bond dipoles in organic molecules is a great aid

.1?I 0
Pulled in this
to predicting chemical reactivity. Species with partial or full negative charges should be at-
tracted to the 8+ region in a molecule or the positive end of the bond dipole. Conversely,
positively charged species would be attracted to the 8- region of the molecule or the nega-
tive end of the bond dipole (see Chapter 10 for the use of these guidelines in predicting re-
H / C'

activity). Such attractions are crucial in controlling weak, noncovalent interactions such as
solvation and molecular recognition (see Chapters 3 and 4). Just like electrostatic potential
surfaces, molecular and bond dipoles reflect ground states. While the polarization patterns
H--...c/x described here can provide valuable clues to reactivity, it is also crucial to consider how bond
Cl ~--~H polarity affects transition states when discussing reactivity.
1 I
H ___ C ~''''H
Opposite ends
of dipoles attract
Molecular Dipole Moments
Whil e the bond dipoles we have just described are, in a sense, conceptualizations, re-
lated to our notions of electronegativity, the molecular dipole is a well-defined, intrinsic
property of a molecule. A molecule has a dipole moment whenever the center of positive
charge in the mol ecule is not coincident with the center of nega tive charge. This separation
of charged cen ters feeds into Eq. 1.2 also, making it possible to calculate the molecular di-
pole moment.
Table 1.3 lists experimentally determined dipole moments for select molecul es. The
numbers tell the relative separation of charges within the m olecules, thereby giving an idea
of the intensity of the electric field around the molecule. They also give a sense as to how

Molecular Dipole Values*
Compound Molecular dipole (D) Compound Molecular dipole (D)
CCI4 0.0 CH3COCH3 2.9
CHCI3 1.0 CH 3COOH 1.7
CH2Cl2 1.6 CH 3COCI 2.7
CH3Cl 1.9 CH3COOCH 3 1.7
CH 3 F 1.8 C6HsCl 1.8
CH3Br 1.8 C6HsN02 4.0
CH 3 I 1.6 1-Butene 0.34
CH30H 1.7 1-Propyne 0.80
CH30 CH 3 1.3 cis-2-Butene 0.25
CH3CN 4.0 cis-1,2-Dichloroethene 1.9
CH3N02 3.4 Tetrahydrofuran 1.6
CH3NH2 1.3 Water 1.8
' Handbook of Chemistry and Physics, CRC Press, Inc., Boca Raton, FL (1979).

strongly an approaching molecule or charge can differentiate one end of the molecule from
the other or, alternatively, how favorable a potential electrostatic interaction can be. For ex-
ample, all approaches to a molecule wi th a molecular dipole of zero, such as tetrachloro-
methane, encounter essentially the same elec tric field. This argum ent ignores both polariza-
tion effects and higher moments such as quadrupoles. In contrast, the electric fi eld felt by a
molecule approaching a structure with a dipole of 4.0, such as acetonitrile, is quite different,
depending upon the direction of approach. Note that the electrostatic potential surfaces
of several small molecules in Appendi x 2 provide a clear way to visualize molecular dipole
An often informative exercise is to analyze a mol ecular dipole as a vector sum of bond Electric field around CC1 4
dipoles. Exa mples of this analysis are shown in Figure 1.4. Note that in high symmetry cases 80 88
all the local bond dipoles cancel and th e avera II molecule has no molecular dipole. Thus, the

absence of a molecular dipole does not ru le ou t the ex istence of bond dipol es, and the pres- 80 88
ence of bond dipoles does not g uarantee the existence of a molecular dipole.
Several trends emerge from examining Table 1.3. The m ore chlorines attached to meth- 80 88
ane, from CH3 Cl to CC1 41 the lower the dipole. This trend might at first seem counterintu-
itive, because we are progressively adding polar bonds to th e system . However, it can be 88
understood as a consequence of vector mathematics, in which the individual bond dipoles Electric field around acetonitrile
increasingly cancel as the number of chlorines increases. The incorporation of nitro or cyano
groups into molecules results in very large molecular dipoles wh en there are no other bond
dipoles to can cel them. An important feature of dipole moments is illustrated by the fact that
the dipole moments of CH3 Br and CH3F are the sa me. We would expect a much larger charge
polariza tion in the C-F bond compared to the C-Br bond, and thi s is so. However, the C-Br
bond is longer th an the C-F bond, and even though the charge separa tion is smaller, the dis-

tance is la rger. The two phenomena both affect the molecul ar dipole, and coincidentally lead
to the sam e dipole moment for the two compounds.

Molecular Quadrupole Moments
Cl ~H
Cl ~CI

In a complete description of a molecule's charge distribution, th e dipole moment is just
one term in a series: monopole, dipole, quadrupole, octupole, hexadeca pole, etc. A mono-
Cl molecular
pole is just a point charge-the dominant term for ions. For neutral molecules organic chem- 1 dipole

ists usually trunca te the series after the dipole. However, the quadrupole moment of a mole-
cule can often be quite important. As such, we take a moment here to remind you about some No molecular
basic electrostati cs. dipole
A quadrupole is simply two dipoles aligned in such a way that there is no net dipol e
(if there wa s a dipole, we' d have a dipole, not a quadrupole). Interestingly, the multi pole + - - Bond
expansion follows a familiar topo logica l pattern. Monopoles look likes orbitals (spheres); +-- Molecu lar
dipoles look like p orbitals (a +end and a- end); quadrupoles look li ked orbitals; octupoles dipoles
like f orbitals, etc. The analogy between multi poles and orbitals is given just to illustrate
Figure 1.4
phasing properties; orbitals do not have polar character.
Examples of analyzing
Figure 1.5 illustrates this point. The most common quadrupole has the two dipoles side- m o l ecul ar dipol es as a
by-sid e pointing in opposite directions, giving four charge regions (two + and two - )and sum of bond dipoles.
the topology of a d xy orbital. This topology is also the arrangement in a quadrupol e mass
spectrometer. However, there is an alterna ti ve arrangement-tw o end-to-end dipoles point-

Monopoles look Dipoles look Quadrupoles look Figure 1.5
like s orbitals like p orbitals like d orbitals The topol ogica l relationships b etween atomic
orbital s and electros tatic moments, and the

~ The quadrupole moment of benzene,
quadrupole m oment of benzene.

~ viewing the molecule edge on


ing in opposite directions-the topology of a dz2orbital. Actually, this arrangement is more
important in organic chemistry, because it is present in benzene (see below). The multipole
exp an sion series-monopole, dipole, quadrupole, etc.-is not a perturbation series. It is not
true that quadrupoles are somehow intrinsically weaker than dipoles in electrostatic inter-
actions. In fact, in some important organic molecular recognition phenomena, quadrupoles
prove to be stronger than dipoles (see Ch ap ter 3).
The most common and important quadrupole moment in orga nic chemistry is that of
benzene. Experimental measurements have determined that benzene h as a large quadru-
pole moment, with a charge distribution as in Figure 1.5 (see Appendix 2). Just as with mo-
lecular dipole moments, we can rationalize a molecular quadrupole moment as a sum of
bond dipoles. In this case, we add six c <>--H8 + dipoles to get the molecular quadrupole. The
exis tence of a large, permanent quadrupole moment in benzene is unambiguous proof that an
sp 2 Cis more electronegative than H. We must have six c <>--HD+ dipoles to explain the effect.
Note that cyclohexane has a negligible quadrupole moment, indicating that an sp3 C and H
have similar electronegativities.

1.1.10 Resonance
The bonding model we have developed thus far is quite "classica l", relying on fairly
simple notions, su ch as Lewis structures. Some structures, however, cannot be adequately
d escribed by a single Lewis structure. In these cases, two or more Lewis structures are
d rawn, and the actual m olecule is a h ybrid or mixture of these resonance structures. The su-
perposition of two or more Lewis structures to describe the bonding in a molecule is ca lled
resonance (also known as mesomerism in very old literature).
A classic example of resonance occurs for acetate. Two structures showing different po-
sitions for the double bond and the negative charge are possible. In thi s case the two struc-
tures are identical, and the charge on each oxygen is -~- Another familiar case is benzene,

0- 0 which also involves two equivalent s tructures, so that the C-C bond length is appropriate
for a bond order of 1.5. All six bonds are equivalent and are represented by two equ ivalent
reson ance structures (called Kekule structures).
Kekule structures

Acetate resonance structures

Resonance structures are not separate molecules that are interconverting. There is really
only one structure, which is best thought of as a hybrid of the various resonance structures.
The two C-0 bonds of aceta te are equivalent and the negative charge is distributed eq ually
between the two oxygens. Often, one symbolizes a combination of the resonance structures
by a single structure meant to describe the hybrid.
Although the examples of acetate and benzene are ones in which the resonance s truc-

-- .. / N, ..
e·.Q. e Q:e
tures are equi valent, this is not usu ally the case. For example, p-ni trophenol and methylvi-
nylketone also have reasonable resonan ce structures, but they are significantly different in
the arrangement of the bonds, lone pairs, and charges.
The picture of resonance implies that the electrons are covering a larger number of
atoms than given by an y one resonance structure, and this is defined as delocalization. Gen-
.. 8 erally sp eaking, the more resonance s tructures that a molecule has and the more reasonable

-- ("
Acceptable resonance structures
these structures are, the more stable the molecule. The energy of stabilization impa rted by
resonance is called the resonance energy or delocalization energy. The reason that a mole-
cule w ith resonance structures is considered more stable is the effect of delocalization. As we
will see in Chapter 14, the more spread out the orbital that electrons occupy, the lower the en-
ergy of those electrons. This is related to a calcu lation often covered in physical chemistry
classes, called the "particle-in-a-box" calculation, w hich we briefly review in the following
Going Deeper highlight.


Going Deeper

Particle in a Box region s of space. The lesson is that if the electrons are
The ma nner in which one finds the energy of electrons is allowed to occupy a la rger amount of space, their energy
to solve the Schrodinger equation (H'Jl = E'Jl). As a ve ry decreases. Specifically, the ki netic energy of th e electrons
simple example, imagine an electron in a one-dimen s io nal drops, which will be a key issue we discu ss in Cha pter
"box" . Here, the potential ene rgy of the electron is zero if 14. Because resonance yie lds a picture of bonding that
the electron is wi thin the box, but is infini te at the ed ges spread s the electrons ou t in space, it is a stabilizing
and beyond the ed ges of the box. The potential energy concept.
cannot be infinite, so the electron is confined to the reg ion
within the box. The solutions(£, ) to the Schrodi nger equa- 00

tion for this scenario a re very simple, and take the form
shown (consult a ny undergraduate physical chemistry
textbook to see h ow the solutio ns are derived). The possi- P.E.
ble ene rgies are qu antized (n is an integer, 1, 2, 3, ... ), w ith L
the le ng th of the box (L) in the denominato r (111 is the m ass
of the electron and his Planck's cons ta nt). As the box gets Oc__ _ ___,
bigger, the ene rgy decreases.
The " box" is an analogy to an orbi tal. With an o rbita l Parameters for the particle in a box
the electrons have their greatest probabili ty in certain

In order to consider what are appropriate resonance structures for a molecule, we first
draw all the possible Lewis structures. In these structures only electrons are allowed to
" move around" . The positions of the nuclei never change. The Lewis structures can have the
maximum number of electrons appropriate for each atom (for exa mple, eigh t for second-
row atoms) or fewer electrons. Next, judgment mus t be made about w hich resonance struc-
tures are reasonable, and Figure 1.6 gives some guidance. Factors that contribute to making
a particular resonance structure acceptable include having a noble gas configura tion for the
atoms, a maximum number of covalent bonds, a m inim um number of li ke charges, close
proximity of unlike charges, and placem ent of negative charges on electronegative a toms.
Not all these guid elines need to be m et to m ake a reasonable resonance structure. Many rea-
sonable stru ctures may contribute only a little to the true electronic structure of the m ole-
cule, depending upon just how reasonable they are. Yet, the identification of all reasonable
resonance structures imparts iniorma tion about polari ty and polarizations in a molecule. As
already mentioned, it is generally true that the larger the number of reasonable resonance
structures associated with a molecule, the more stable it is. In addition, resonan ce is espe-
cially favorabl e when it involves two or more equivalent resonance structures (as, for exam-
ple, with acetate and benzene), and when all second-row atom s have a full complement of
eight valence electrons.

I .
.. /·· vy ·
N ·a

:QH : QH

6 .. 6
Reasonable resonance structures Unreasonable resonance structures in shading

Reasonable and unreasonable resonance structu res.


N' N° Ne N° Ne 8_
Oo -- ~
.. 8 - - · - - -8 .. /; - - o·-

Resonance structures of pyrrole

Often we will see an atom adopt a nonstandard h ybridization in order to maximize reso-
nance. This most commonly arises when an atom h as a lone pair of electrons that is in conju-
gation w ith (directly b onded to) a TI system . For example, consider pyrrole, shown above.
The nitrogen atom would seem at first glance to b est be described as an sp 3 h ybrid, b ecause
four group s are attached to it: two N-C bonds, an N-H bond, and a lone pair. However, in
order to accommodate the reson an ce structures sh own, four of which have a double bond
between Nand a C, the lone pair must be in a p orbita l, not an sp 3 hybrid orbital. This require-
ment makes the nitrogen atom sp 2 h yb ridized, which is experimentally supported by the fa ct
that the N is trigonal planar. Such resonance effects on hybridization are common and
should be routinely looked for when assigning hybridization to various atoms. Another ex-
ample is the h ybridization of N in an amide, which sh ould be classified as sp2 . However, as
shown in the foll owin g Connections highlight, the validity of resonance in an amide has re-
cently come into question .

1.1.11 Bond Lengths
The simple bonding model we have d eveloped thus far can rationalize many geometri-
cal features. We h ave already discussed bond angles, and m ost organic chemists know what
standard angles are for organic structures. It is genera lly true tha t it is more difficult to dis-
tort bond lengths from standard values than it is to b end angles. Nevertheless, it is worth-
while to know som e standard bond len gths, as a significant deviation from these values is a
clear indication of s ubstantial strain in a m olecule or som e non-standard bonding situation
(see Chapter 2). In addition, we need to know bond lengths to evaluate bond and molecu-
lar d ipoles.
Several trends can be gleaned from the series of average bond lengths listed in Table 1.4.

Table 1.4
Typical Bond Lengths of Some Covalent Bonds*
Bond Length C.\.) Bond Length <A)
Single bonds Double bonds
C(sp3 )-C(sp3) 1.53-1.55 C(sp2 )- C(sp2 ) alkenes 1.31-1.34
C(sp3 )-C(sp2 ) 1.49-1.52 C(sp2 )-C(sp2 ) arenes 1.38-1.40
C(sp 2)- C(sp 2) conjugated 1.45-1.46 C(sp2)-0 (sp2) aldehydes and ke tones 1.19-1.22
nonconjugated 1.47- 1.48 C(sp2 )-0(sp2 ) esters 1.19-1.20
C(sp )-C(sp) 1.37-1.38 C(sp2 )-0(sp2 ) amides 1.225-1.24
C(sp3 )-0(sp3) ethers 1.42-1.44 C(sp2 )-N (sp2 ) imines 1.35
C(sp3 )-N (sp3 ) amines 1.46- 1.48 Triple bonds
C-F 1.39- 1.43
C(sp )- C(sp) alkynes 1.17- 1.20
C-CI 1.78- 1.85
C- Br 1.95-1.98
C-1 2.15-2.18
C(sp3 )-H 1.09-1.10
C(sp 2 )-H 1.075-1.085
C(sp)-H 1.06
N-H 1.00- 1.02
0 -H 0.96-0.97

*Allen, F. H., eta/. "Tables of Bond Lengths Determined by X-ray and Neutron Diffraction. Part 1. Bo nd Lengths in Organic Compounds."}. Cin'llt. Soc.,
Perkin Trans. II, S1- Sl9 (1987).



Resonance in the Peptide Amide Bond? More recen t w ork, however, has suggested a refine-
A found ation of structural biology and bioorganic chem- ment ofth e traditional model. Modern computational
istry is the peptide bond, the link between consecutive methods allow detailed analyses of molecular and elec-
amino acids in a protein. Interestingly, the modern view tronic s tructure . These tools were used to look in detail
of this p rototype structure is evolving, and as such it pro- at amide rotation. The classica l resonance model predicts
vides an excellent example of various rational izations of C-N double-bond character and C-0 single-bond char-
structure and bonding. acter in a p la nar amide. Since resonance is destroyed on
The peptide bond is an amide formed between a pri- rota ting 90° about the C-N bond, we would exp ect the
mary amine and a carboxylic acid. The most important C-N bond to leng then and the C-0 bond to shorten in the
structural features a re (a) the amide group is pl ana1~ with pe rpendicular form. Calcu lations do show that the C-N
a substa ntial barrier to rotation (typically in the range of bond leng thens upon rotation, but the C-0 bond length is
~ G* = 15-20 kcal I m ol); and (b) there is a significant prefer- essentiall y unch anged. Also, examination of the charges
ence for the Z confo rmer (termed " trans" in the protein lit- calculated on the atoms of a typical amide do not suppo rt
era ture), which places theN-H and C = O bo nds trans to structure B. TheN of an amide is no t significa ntly more
each other. Ano ther key feature of ami des is their strong posi tive than theN of an amine, for which such resonance
propensity toward h ydrogen bonding. Two key protein is not possible. Also, the 0 of an amide is not significantly
secondary structural elements- the a-helix and the more nega tive than the 0 of a ketone. Finally, the C of an
13-sh eet- depend on the amide group to act as both a amide is quite positive, just like the C of a ketone.
hydrogen bond donor and a h ydrogen bond acceptor. These observa tions suggested that the simple two-
structure resonance model was inadequ ate. Instead three
0 structures seem necessary, as sh own. Both resonan ce struc-
~Gt = 15- 20 kcal/mol R/ C,N/ H
tures B and C are considered to be of major importance for
I the rotation barrier. In this view, the role of the oxygen is
to polarize the C-0 bond, introducing a large o+ on car-
z E bon (resonance structure C). TheN d oes n ot develop a
(trans) (cis)
significant 8+, despite the implications of reson ance
R' structure B. Ins tead, this model em ph asizes dipole inter-
R'\ 'N-
H ··· actions- the C- 0 dipole is aligned for a favorable inter-
R' N- H···O =C action with the N- H bond dipole. Note that a o+ on N
' N- H ··· O=C' 'R
'R would not be consistent with this p icture.

Features of am ides

In the traditional m odel of an amide, resonance is the
key concept. As shown below, one can w rite a reason able
Amide bond dipoles
resonance structure for a n amide tha t places a double
bond between the C and theN (structu re B). This " double-
bond ch aracter" lead s to a planar structu re, and hindered This refi ned resonance model was inspired by high-
rota tion about the C- N bond. To the extent that electro- level calculatio ns, and it re fl ects chan ging vie ws on the
sta tic interaction s contro l hydrogen bond s trengths (see importance of bond dipoles (i .e., CO+ -O<>-) in structure and
Ch apter 3), the ch arges implied by reson an ce structure B reactivity. Bond dipoles are also important in rationalizing
su ggest strong hydrogen bonding in amides, as is the Z- E preference. There are two strong bond dipoles in a
observed. secondary amide: CO+ -as- a nd H<>+ -N,_. As shown, the Z
form aligns these two dipoles favorab ly (the positive end
near the negative end), but theE form would produce an

-- B.
-- c.
unfavorable alignment. Finally, the strong propensity of
the peptide bond toward hydrogen bonding is readily
understood based on the charge distribution implied
by the resonan ce and bond dipole argumen ts.
Wibe rg, K. B. "The Interaction of Ca rbonyl Grou ps with Subs tituents."
Amide resonance structures
Ace. Chem. Res., 32, 922- 929 (1999).


One obvious trend is that multiple bonds tend to be shorter than single bonds. The bond
lengths in arenes are between C-C and C = C bond lengths, just as predicted by resonance
theory. Bonds involving atoms with larger atomic radii (further down in a column and / or to
the left within a row in the periodic table) are longer.
The state of hybridization affects bond lengths; mores character in the hybrid decreases
the bond length. This effect is another manifesta tion of the fact that s orbitals lie closer to the
nucleus than p orbitals. This last effect is sometimes under appreciated. For example, the fact
that the C2-C3 bond of 1,3-butadiene is much shorter than a typical single bond in an alkane,
1.48 Avs. 1.54 A, might lead to the conclusion that there is some double-bond character in
C2-C3 (see the resonance structure in the margin, which is sometimes taught in introduc-
Butadiene resonance structures
tory organic chemistry). However, Table 1.4 shows that this is primarily a hybridi zation ef-
fect, as system s that are not planar (nonconjugated and hence have no double-bond charac-
ter) but are sp 2- sp2 single bonds, show comparably short bond lengths. Further support that
the conjugation in butadiene has only a small effect on the C2-C3 bond length comes from
the rotation barrier about this bond, which is only on the order of 4-5 kcal I mol (see Chapter
2 for an expanded discussion of rotational barriers).
Bond lengths can also be understood w ith reference to the radii of the constituent atoms.
There are actually three different kinds of radii that we use to understand molecular dimen-
sions: covalent, ionic, and van der Waals. The covalent radius is half the dis tance between
two identical atoms bonded together. For example, the C-C bond length in ethane is 1.54 A,
and therefore the covalent radius of C is 0.77 A. The ionic radius of an atom is its size as deter-
mined in the crystal lattice of various salts. The radius d epends upon the charge on the ion,
and is sm aller as the positive charge increases and larger as the nega tive charge increases. Fi-
nally, the van der Waals radius of an atom is the effective size of the electron cloud around an
atom when in a covalent bond, as perceived by an atom to which it is not attached . The mea-
surem ent comes from an analysis of crystal packing, and effectively sets the steric size of an
atom or a group. Table 1.5 shows just a few covalent, ionic, and van der Waals radii. There is
som e debate as to the exact values of some of these, especially the van der Waals rad ii, but
the basic trends are clear. Using these values molecular surface areas and m olecular volumes
can be calculated.

Table 1.5
Covalent, Ionic, and van der Waals Rad ii of Select Atoms (A) *
Atom Covalent v ow Ion Ionic
c 0.77 1.68
H 0.30 1.11 H- 2.08
N 0.70 1.53
0 0.66 1.50
F 0.64 1.51 p- 1.36
Cl 0.99 1.84 Cl- 1.81
Br 1.14 1.96 Br- 1.95
I 1.33 2.13 I- 2.16
*Pau ling, L. (1960). Tl1c Nntureofthe Che111icnl Bond n11d the Structure of Molecules
a11d Crystals; nn Int roduction to Moder11 Structural Che111istry, 3d ed., Corne ll Univer-
sity Press, Ithaca, NY.

1.1.12 Polarizability
An important property of molecules that we have yet to di scuss is polarizability. Since
electrons are charged parti cles they respond to electric fields. In particu Jar, electrons in m ole-
cules are mobile to varying degrees, and so their positions can sh ift to diffe ring extents in re-
sponse to an applied electric fi eld. The electron cloud is said to become polarized in response
to the electric field. The ability of the electron cloud to distort in response to an external field
is known as its polarizability. Upon distortion, a dipole is typically induced in the molecule,
adding to any permanent dipole already present.

1 . 1 A R E VI E W O F BAS IC BO Dl G CO C EPTS 25
Table 1.6
Atomic and Molecular Polarizabilities (a, cm 3 /10-24 )*
Atomic polarizabilities
H 0.6668 He 0.205
c 1.76 N 1.10 0 0.802 F 0.557
p 3.13 s 2.90 Cl 2.18
Br 3.05
I 4.7 (or 5.35)
Selected molecular polarizabilities
CH4 2.6 NH3 2.21 H zO 1.45 H 2S 3.8
COz 2.91 CSz 8.8 CF4 3.84 CCI4 11.2
CzH z 3.6 CzH " 4.25 C2H6 4.45 CH 30H 3.23
Benze ne 10.32 Cyclohexene 10.7 Cyclohexane 11.0
*CRC l-ln11dbook ofCil emisfry n11d Physics, D. R. Lid e (ed .), C RC Press, Inc., Boca Ra to n, FL (1990- 1), pp. 10-193-10-209.

We define the polarizability of a molecule as the magnitud e of the dipole induced by one
unit of field gradient, which works out to be in units of volume. Often, the larger the volume
occupied by the electrons, the more polarizable those electrons. Since an electric fi eld gradi-
ent is directional, it can encounter a molecule or bond in different ways depending upon the
relative orientation between the field and the molecule or bond. Hence, polarizabilities are
often broken down into one longitudinal (along the bond or molecular axis) and two trans-
verse (perpendicular to the axis) values. Table 1.6lists some atomic and molecular polariz-
abilities, for which the directional components have been averaged. Just as with permanent
dipoles, the induced dipoles that arise from such polarizations can be analyzed as molecular
dipoles or as bond dipoles. We can thus speak of molecular I atomic polarizabilities, as well
as of bond polari zabilities.
Several trends are evident in Table 1.6. First, as we move left to right across a row of the
periodic table, polarizability decreases. This trend is clear in atomic polarizabilities (C > N
> 0 > F) and molecular polarizabilities (CH4 > NH3 > H 2 0). Electronegativity plays an im-
portant role. Atoms that hold on to their electrons tightly are not polarizable.
Polarizability has profound consequences. Water is a very polar molecule-it has a large
dipole moment. However, methane is much more polarizable than water, and alkanes in
genera l are among the most polarizable of molecules. Thus, while aqueous media provide a
very polar environment, alkanes and other hydrophobic environments are more polarizable.
The second clear trend in Table 1.6 is that as we move down a column in the periodic ta-
ble, polari zability increases substantially (S > 0, P > N, and H 2S > H 20). This observation
explains an earlier apparent contradiction. According to Table 1.1, I is not significantly more
electronegative than C, yet C-I bonds are much more reactive than C-CI bonds in reactions
like SN2 and E2. In such instances, C8+-x<>- polar bonds are often invoked, but C-I bond s are
not very polar. However, dramatic changes in molecular and electronic structure occur in
the course of a reaction. A model for just the ground state of the reactantis often inadequate.
When an anionic nucleophile approaches a C-X bond for an SN2 reaction, it can induce a
large bond dipole, especially if X is highly polarizable. Thus, the large polarizability of I
makes up for the lower electronegativity, and the C-I bond is more reactive than C-CI.
Another perhaps surprising point is revealed when considering the polarizabilities of
the simple hydrocarbons given in Table 1.6. Because of their greater reactivity, most chemists
might assume that alkenes are more polarizable than alkanes, but the opposite is true. Once
again, alkanes are among the mos t polari zable of molecules. In particular, ethane is signifi-
cantly more polarizable than ethylene. Perhaps even more surprisingly, cyclohexane is more
polarizable than benzene. While perhaps counterintuitive, these trends are consistent with
electronegati vity arguments. Alkenes and arenes, with the more electronegative sp 2 carbons,
are less polarizable than alkanes with only sp 3 carbons.


1.1.13 Summary of Concepts Used for the Simplest Model
of Bonding in Organic Structures
The material presented to this point has been mostly a review of concepts presented in
introductory organic chemistry classes. Only in a few cases was the discussion extended fur-
ther. This review sets the stage for us to go deeper into bonding and its effects on structure,
and a d eeper analysis is what we present in the remainder of this chapter. First, however, it
is in structive .o summarize the concepts used in the simplest picture of bonding in organic
structures; many of these concepts extend nicely into our second approach to bonding.

Hybridization. Atoms contribute consistent sets of hybrid atomic orbitals to the
bonding in a molecule. A carbon with four ligands is sp 3 h ybridized, a C with three li gands is
sp 2 hybridi zed, and a C with two li gands is sp hybridi zed .

cr and 1t Bonds. Hybrid orbitals on two separate atoms overlap in the region between
the adjacent atoms to create a bonds. The overlap of p orbitals on adjacent atoms creates 7T
bonds. The localized bonds consist of bonding and antibonding molecular orbitals, of which
only the bonding orbitals are populated with electrons. This arrangement of bonds is a VBT
notion, with discrete and localized bonds between adjacent atoms, but the orbitals are cre-
ated using MOT mixing notions.

Resonance. For certain molecules, a single valence bond stru cture cannot properly
describe the bonding, so resonance is involved. Resonance struch1res show various arrange-
ments of electrons within a structure, where each contributes to th e bonding arrangement in
that molecule. Resonance structures can also be used to s uggest subtle features of the elec-
tronic s tructure of functional groups.

Electronegativity and bond polarization. Electronegative elements pull electron den-
sity toward themselves. This introduces polarity into bonds, res ulting in bond dipoles and
mol ecular dipoles.

Induction. Electronegativity and polarization effects can be felt a few bonds away
throug h an inductive effect.

Polarizability. Polarizability reflects the extent to which electrons can be perturbed
from the s tandard bonding arrangement in response to the approach of another m o lecu le.
This effect is important in unde rstanding solvent properties and many reactivity patterns.

Sterics. Two or more groups tend to stay away from each other due to adverse electro-
static repulsion between filled orbitals.

We draw upon each of these notions as necessary to explain various aspects of bonding,
reactivity, and structure. We will add solvation effects later in the book. For parti cular reac-
tion s or s tructures, we may have to refine and/ or combine th ese notions to ge t the optimal
model for the molecule. In some cases we may even need to completely re-think and modify
these foundations. However, there is often a tendency in physical organic chemistry to be-
come quite focused on the "exce ptions to the rule" . We should keep in mind that the vast ma-
jority of organic structures are well d esc ribed by this simple model.

1.2 A More Modern Theory of Organic Bonding

In Section 1.1 we described the basic bonding patterns of organic molecules and the proper-
ties of different types of localized bonds. The concepts introduced were mostly fairly classi-
cal valence bond concepts. They have definite predictive power, and that is an important
measure of the value of any model. However, the picture we are about to present also has the


same predictive power, but can better explain certain structural issues and experimental ob-
servations. In addition, understanding chemical reactivity poses a serious challenge, and
this second approach to bonding is very good at predicting the reactivity patterns of organic
The highest extent of rigor that can be given for bonding is presented in the advanced
quanh1m mechanical analysis that was developed in the first half of the 20th century. Mod-
ern calculational methods now provide accurate representations of the molecular orbitals
not only of stable molecules, but also of reactive intermediates and even transition states.
These powerful computational methods will be described in detail in Chapter 14. Our goal
in this chapter is to develop an understanding of chemical bonding, and the detailed numeri-
cal output of a quantum mechanjcal calculation is not enough. Here, we will show that cer-
tain key concepts and trends that result from the output of such calculations lead to a more
rigorous descriptive model of organic bonding than given in the first part of this chapter.
Molecular orbital theory (MOT) forms the core of this second model. However, we will see
that certain key concepts from the valence bond theory based model for bonding will still be
very useful (e.g., sterics, induction, and polarizability), and in fact the distinctions between
the two theories are often blurred. Therefore, we again will present a hybrid VBT /MOT
model, but now more tilted toward MOT.
Although the starting points for VBT a nd MOT are different, and the initial mathematics
used to explain them are different, both theories can be shown to give similar results when
correction factors are added and their respective mathematics are solved completely in a
manner consistent with quantum mechanical rules. In addition, straightforward and theo-
retically justifiable mathematical operations can, in most cases, transform the fully or par-
tially delocalized molecular orbitals we are about to examine into localized orbitals that
clearly resemble our VBT notions of discrete and localized bonds. MOT and VBT are not as
far apart as they may seem. Hence, one theory is not necessarily more correct than the other.
Also, there is nothing wrong with a hybrid approach to bonding. Both MOT and VBT are ap-
proximations to the "true" answer-a full solution to the Schrodinger equation. A combina-
tion of the two is no less approximate, as long as it is thoughtfully applied.

1.2.1 Molecular Orbital Theory
In contrast to VBT, "full-blown" MOT considers the electrons in molecules to occupy
molecular orbitals that are formed by linear combinations (addition and subtraction) of all
the atornic orbitals on all the atoms in the structure. In MOT, electrons are not confined to an
individual atom plus the bonding region with another atom. Instead, electrons are con-
tained in MOs that are highly delocalized-spread across the en tire molecule. MOT does not
create discrete and localized bonds between neighboring atoms. An immediate benefit of
MOT over VBT is its treatment of conjugated 1T systems. We don't need a "patch" like reso-
nance to exp lain the structure of a carboxylate anion or of benzene; it falls naturally out of the
delocalized nature of the MOs. The MO models of simple molecules like ethylene or formal-
dehyde also lead to bonding concepts that are pervasive in organic chemistry.
On the other hand, the fully delocalized view of MOT that is so useful for certain mole-
cules like benzene is not so useful in other organic molecules. For example, if we want to an-
ticipate the reactivity of 3-heptanone, we really don't need to consider orbitals that span all
seven carbons plus the oxygen. Chemical experience tells us that the action is at the car-
bonyl, and we want to be able to focu s on the 1T system and lone pairs. In order to do so, we
will use an MO approach to establish certain fundamental bonding principles by studying
small prototype molecules-ethylene for learning about C=C bonds, formaldehyde for
C = O, acetic acid for carboxylic acids, etc. This analysis gives us the MOs for these functional
groups. Fortunately, analysis of the computed MOs of complex systems shows that the MOs
of the smaller model molecules appear with only minor modifications in the larger mole-
cules. In other words, we can concentrate our analysis on the molecular orbitals of functional
groups to understand structure and reactivity. Thus, a key concept we present below is that
of group orbitals-orbitals that are delocalized only over a defined group of atoms. If we
need to understand the molecular orbitals of the entire molecule, we combine these group


orbitals to create molecular orbitals for the entire molecule, and the rules for this orbital mix-
ing procedure are presented in the following discussions.
To create group orbitals or d elocalized molecular orbitals, we need to understand how
to combine atomic orbitals properly. Therefore, the starting point in developing our second
model of organic bonding is a set of rules that lead by inspection to group orbitals and molec-
ular orbitals. This procedure is called qualitative molecular orbital theory (QMOT).

1.2.2 A Method for QMOT
The protocol we will follow was developed by many workers, including Hoffmann and
Salem. An especially succinct statement of the procedure embodied in a series of rules was
provided by Gimarc, and is presented in Table 1.7. Below we give several examples of using
rul es 1- 13, where rules 14 and 15 become more important in future chapters. Several MOT
and quantum mechanical concepts go into the origin of these rules. For example, the orbital
mixing follows the well-developed rules of perturbation theory, which we will often discuss
(glance back to Section 1.1.7 for our first quick look at mi xing). Furthermore, the symmetry
of the molecule (see the next Connections highlight) guid es the crea tion of the MOs, a no tion
that h as a quantum m echanical origin discussed in Chapters 14 and 15.

Table 1.7
The Rules of QMOT*

1. Conside r valence orbitals o nly.
2. Form completely deloca lized MOs as linear combinations of sand p AOs.
3. MOs must be ei ther symmetric or antisy mme tric w ith respect to the symmetry operations of the molecule.
4. Compose MOs for stru ctures of hig h symme try and then produce orbitals for re la ted but less symmetric
structures by systematic distortio ns of the orbitals for higher symme try.
5. Molecules with simila r molecular structures, s uch as CH 3 and H 3, have qualita ti vely s imilar MOs, the
major difference being the number of valence electrons tha t occupy the commo n MO system.
6. The total e ne rgy is the sum of the molecular orbita l energies of individual valence electrons.
7. If the two hig hest energy MOs of a given symmetry derive primarily from diffe rent k inds of AOs, then mi x
the two MOs to form hybrid orbi tals.
8. When two orbitals interact, the lower energy orbital is stabilized and the higher energy orbital is
destabi lized. The ou t-of-phase or anti bonding interaction between the two sta rting o rbitals always ra ises
the energy m ore than the corresponding in-phase or bonding interaction lowers the e nergy.
9. When two orbitals interact, the lower energy orbital mixes into itself the higher ene rgy one in a bonding
way, whi le the higher energy orbital mixes into itself the lower ene rgy one in an antibonding way.
10. The smalle r the initial ene rgy gap between two interacting orbita ls, the stronge r the mixing interaction.
11. The larger the overlap be tween interacting orbita ls, the larger the interaction.
12. The more electronegative elem ents have lower e nergy AOs.
13. A change in the geometry o f a molecule wi ll produce a large ch ange in the energy of a p a rticular MO if the
geometry ch ange resu lts in changes in AO overlap tha t are large.
14. The AO coefficients a re la rge in h igh ene rgy MOs w ith ma ny nodes or complicated nodal surfaces.
15. Energies of orbitals of the sam e symm etry classification cannot cross each other. Instead, such orbitals mi x
and dive rge.

*Adapted, w ith m odifications, from Gimarc, B. M. (1979). Molecular Structure and Bonding: Tile Qualitative Molecular Orbital Approach,
Academic Press, New York.

. role in QMOT because the concept is used in rules 3.. we form the delocal- ized MOs using the other rules. I H'""·· M-H H. The starting point is planar CH3. with the examples given below. an o rbital must be symme tric or antisymmetric. With respect to each symmetry operation.. To begin with. Wiley the orbita Is. we will look one we are going to need until Chapter 15. and given in the end-of-chapter Exercises. This structure possesses one C3 axis. a carbon 2s orbital. and their relationships to orbitals is beyond the their meaning in context is usually clear. To under. the 0 311 scop e of this book (see the reference at the end of this high. H ligands belongs to the point group ca lled 0 311 .. we w ill not worry as to w h ether we have methyl cat- ion._ A trigonal planar stru cture with three equivalent H. For an orbital to be symmetric..7 that symmetry plays a n important the orbital must be reversed by the symmetry operation. returning exactly the same structure with atoms returned to ide ntical positions in space. a bond (only one is shown).. We are choosing a reactive species such as this. f01~ as we w ill describe later. w h ere reflection of the atoms throu gh these mirror p lanes likewise g ives back the same structure. because we want a simple structure as an easy starting point. respectively. However. Planar Methyl Because of ru le 1. no C2 axes. Che111ical Applications of Group Theory. you should notions of symme try will hopefully become evident be awa re of the symmetries inherent in only two geome. but with a little practice it should become clear. It also pos- sesses three C2 axes perpendicular to the central C3 axis. we discuss these two systems briefly. we'll put the electrons in later.... on ly one is shown). To be antisymmetric..--. While these are not precisely defined terms. we will "build" the MOs of planar CH3 from scratch. structure is of higher symmetry than the C3v . point s tructures". 1 .. N ext. For now. one along each M-H bond vector (only one is shown to the right). and three carbon 2p orbitals. tetrahedra l where one group is diffe rent than the other y three (also called pyramidal). H I":J. 2nd ed.--. For example... 7. le t's just make the MOs of planar CH3 . or radical. but does h ave three <J planes (one d e fined by each Symmetry operations H. the orbitals we have to work with for meth yl are simple: three hydro- gen l s orbitals. A full explanation of symm etry operations. The MH3Y s tructure belongs to the point group called C3. A 0 311 struc- ture a lso possesses three intern al mirror planes. Certainly. the orbital diagram is not o verly sensitive to such distinctions....2 A MORE MODERN THEORY OF ORGAN I C BOND I NG 29 Connections A Brief Look at Symmetry and Symmetry Operations metry operation. Interscience. .. They contain th e C3 axis and lie along each M-H . F.7. Rule 4 speaks of " high symmetry" and "less symmetric and 15. The use of these rules and the stand the QMOT examples discussed below.2. How this is done may at first seem mysterious. To possess a C3 axis and a C2 axis means that the molecule can be rotated along this axis by 120° and 180°. The struc- Trigonal planar 0 3h Pyramidal C3 v ture possesses a c3axis p erpendicul ar to the p lane of the molecule that passes throu g h the central ato m.. Rules 3 and 7 address the symmetry properties of Cotton. A. addressed at that time. anion. called <J planes. and with the p ractice tries: trigonal planar with all three groups the same. groups.M-Y plane. and those are shown in Figure 1.. using . {1971). Hence. 4. so it will be at prope r and im proper rotations in Chapter 6. New York. it mu st be unchanged by the sym- 1. all signs (phases) of It is clear from Table 1. because p yramidal CH3 is of lower symmetry (rule 4).M--. Rule 15 is not light for an excellent discussion).3 Methyl in Detail To illustrate the procedures of QMOT.

so we mix theses orbitals in phase.7.7 we only show five orbitals. as shown in B and C in the figure. please simply accept this degeneracy as fact for now. in Figure 1. C. The Pz AO cannot mix with any of the h ydrogens.7 The orbitals of a planar methyl group.1 AOs can mix with the h ydrogens to give favorable interaction patterns. and this carries over to the MOs. Next come the carbon p orbitals. . meaning that they exist but are not typically populated with electrons (see the dis- cussion of adding electrons below). Our goal here is to fo- cus on the low-lying.7.7. and so Px and py are degenerate. with energy increasing as we go from the bottom of the diagram to the top . In this high symmetry system. The Px and p. If you are uncomfortable with this symmetry argument. We thus have a molecular orbital that is just an atomic p orbital. We do not draw two of them becau se they are not in- volved in bonding schemes. antibonding orbital. There must be a conservation of the number of orbitals. Mixing just means adding and subtracting. as are the MOs derived from them. Orbital A is symmetric with respect to the C3 axis in the molecule. as they all lie on the node of this orbital. meaning that the number of molecu- lar orbitals created must equal the number of atomic orbitals we start with. Yet. and we do this visu- ally. The reason A lies below B and Cis that carbon 2s AOs are lower in energy than 2p AOs. B. Orbitals A-C are stabilized by mixing with the H AOs. Orbital E and the twoMOs that we do not show are called virtual orbitals. ! E H Figure 1. created using the rules of Table 1. cR - H D H 1 6\ H B c ! A rules 2 and 3. we create a high energy. To create the group orbitals for planar methyl we u sed a total of seven atomic orbitals. We wou ld call orbitals A. We also show the relative energies of these MOs in Figure 1. we mix the carbon 2s orbital with the three h ydrogen atomic orbitals (AOs) in- phase to produce orbital A. called a nonbonding orbital. except with excited electronic states. MOs B and Care of equal energy- they are d egenerate. the x andy axes are equiv alent. but Dis n ot. and D the group orbitals for planar methyl. bonding MOs.7. shown as E in Figure 1. If we mix out of phase. and labeled as Din Figure 1.30 C H APTER 1: I NTRODUCTION TO ST RUCTU RE AN D MODELS O F BOND I N G X y J-. in a bonding manner.

A A cr(CH 3 ) Figu re 1. along w ith the results of a calcul ation on pyra- mid al meth yl. On p yramid aliza tion. Figure 1. named after A. the energy of A will chan ge very little. -------- ~ E ~ E' H ~ -- H D H ~ D ::r cr (out) D' H ~ ~ H B i- - c B c n (CH3 ) ~ ~ X ~ --------. As you can see. Also shown are th e com puted MOs for th e system. the fi nal MOs of pyra mida l CH 3 . We make the MOs of a p yramid al methyl group by distortin g the planar system and foll owing the changes as in d ica ted in rule 4. Beginni ng with pl anar CH 3 on th e left and moving ri ght.8. and the conventionaJ labeling of th e gro up orbitals for methyl. Walsh. Deciphering the origin of the calculated orbitals is a major goal of our ana lysis. py ra mid ali za tion lowe rs orbita l A sli ghtl y. You d on't actually see the individu al s and p atomi c orbitals. We w ant our QMOT m od el to faithfull y capture the essen ce of the ac- tu al computed results. but you see the results of mixing them . raises B and C more so.7. it gives us confidence to use the m ethod to pre- dict oth er orbitals. obtained from a qu antum mechanical cal- culation. an d we w ill be constantly checking our qualitative reasoning aga inst quan titati ve calculations to be sure we are gettin g things right.8 shows th e Walsh diagram for CH 3 . our simple qualitative analysis has predi cted th e essence of th e full calculati on qu ite well. the theoretician w ho first devised the approach.2 A MORE MODERN THEORY OF O RGAN I C BOND I NG 31 Figure 1. again using the rules from Table 1. and lowers D substantia ll y. . 1. Also shown is a seco nda ry mixing between D and E to produce D ' a nd E'.7 also shows th e actual MOs of CH3 +. A diagra m tha t follows orbital energies as a functi on of angular distorti ons is called a Walsh diagram. we can predi ct how geometrical di stortions wi ll affec t the MOs. too. and it is rep rodu ced on the left side of Figure 1. D. Perhaps a slight ~ E . they a re not to scale. We begin wi th the plan ar system we just d eveloped.----- )' )-.8 The Wa lsh d iagram for CH 3 . If the QMOT rules prove ef- fective in reprodu ci ng ca lcula ted stru ctures. Atomic orbi tals are drawn to show phase relationsh ips onl y. The W alsh Diagram: Pyramidal Methyl Once we h ave the orbitals for a fragment.

a collecti on of atoms and bonds that shows a consisten t reacti vity pa ttern in a w id e range of molecular environments. The B/C pair w ill rise in energy on p yramid aliza tion significantly more than A is lowered because of the greater directionality of the bonding.1. as shown in Figure 1. because we have not yet considered rule 7. planar methane. We begin with a low-lying. Orbitals A-Care strongly C-H bonding. We will use these group orbit- als to model the bonding in any molecules that contain methyl groups (see Section 1. The two models agree. We are used to thinking of a carbon yl or an olefin as a functional group. whether in the planar or pyramidal forms. jus t as is expected when one creates sp hybrids (th ey point in opposite directions on the sam e atom). but contrary to simple Lewis dia- grams. a collec- tion of partially deloca lized orbitals that is consistently associated with a functional group or simil ar collection of atoms in molecules. C-H bonding orbital d erived from the car- bon 2s orbi tal. Thus. th e group orbitals for methyl are given specific d escriptive names. They are called 1r(CH3 ) (a degenera te pair). To a considerable extent th ese groups will correspond to the familiar functiona l groups. each with two electrons.3). and we will find they play an important role in un- ders tandin g many aspects of organic structure and reactivity. This same notion ca rries over to electron ic stru cture. allowing favorable bonding interactions to develop. A formerly nonbonding or- bital becomes significantly bonding. Next comes another orbital of 0'-type symmetry. Now D ' looks more like a lone pair orbital. we still have three C-H bonds. MOs D an d E have th e same symmetry. it contains contributions from both the 2s and the 2pz atomi c orbitals. They can contain six electrons. possessing no nodes along the z axis (the axis that would bond CH3 to anything else) and is ca ll ed O'(CH3 ). Next comes the C-H bonding orbitals that are derived from carbon 2p orbitals. For conveni ence. The biggest effect by far of pyramidaliza tion is on orbital D. . In the pl anar form the p orbitals can p oint more directly at the hydrogen s. maximizin g overlap. there are group orbitals for these groups. but for a group not normally considered a functional group-methyl. We call th em group orbitals because we can use th ese orbitals to describe any 0' and 1r bond. Certain col- lections of atoms contribute a consistent set of orbitals to any molecule in which they are present.8 shows the result as D' and E'. there is still significa nt C-H bonding in thi s lone pair orbital. However. Let's look at the bonding in the planar and p yra midal forms. and so it is called O'(out). the B/C pair of orbitals remains degenera te.8. and Figure 1. de- scribed by three occupied MOs that are s trongly C-H bonding. but not a methyl or meth ylene group. the anal ysis ju st presented actually defined group orbitals also. one of the goals of introducing a second mod el of bonding is to de- fine th e group orbitals of m any types of functional groups. In a VB m od el. However. Orbital E' also looks like a hybrid. We are not done with the QMOT analysis. The localized bondin g and antibonding C-C 0' and 1r orbitals given in Section 1. the energy of Dis considerably lower in pyramidal vs. "Group Orbitals" for Pyramidal M ethyl One of the m ost important concepts in organic chemistry is th at of the functional group. but there wi ll be some differences. and this destabilizes the orbitals.7 are group orbitals. we would want three C-H bonds. Note also that D' is very much like an sp" h ybrid at carbon. You are alrea dy familiar with a few orbitals that can be considered as g roup orbitals fro m o ur first VBT / MOT model of bondin g. This overlap dimini shes as the sys tem distorts. hence describing any alkane C-C bond an d any C=C fu nctional group. rule 7 tells u s to mix these two orbitals. so it is sensible to consider them to be of 1r symmetry. It points away from the hydrogens. There is a nodal plane that passes through the carbon and contains the z axis in each group orbital of this type. but one is based on a carbon 2p orbital and the other on a 2s orbita l. Even with dis tortion. As mentioned ea rli er. for a total of six C-H bonding electrons.32 C H A PTER 1 : l TRODUCT I O TO ST R UCTURE AND MODELS OF BO N D I C energy lowering will occur because the hydrogens are closer to each other and can interact favorably. It is of 0' symmetry. Thus. but points in the opposite direction from 0 '. and many will be examined as we go along. We will refer to these as group orbitals. With QMOT. The h ydrogen s h ave moved off the node of the p orbital.

and N have th ree. but to an y MH 3. just as in the case with CH 3 . as C-H overlap d rop s. as th e sum of the orbital energies for occupied orbi tals. Since B. four. The actual MOs of th e bent form are also shown in Figure 1. let's briefl y consider only H 20. leav ing a d egene ra te pair of nonbonding p orbitals. seven. where we consider the relationship between the linear and bent forms. Because neutral CH3 is a radical. we have a low-lying.7. On going from lin ear to bent. or NH 3. MO A is slightly stabilized. However.9 shows the Walsh diagram for meth ylene.9. The Walsh Diagram and Group Orbitals Beginnin g w ith th e linea r structure. In contrast. Our sim ple model has thus made a clear prediction: BH3 should be planar. For BH3. CH bonding orbital (A) de- rived from the ca rbon 2s o rbital . Let's foc us first on the even electron systems. while NH 3 should be pyra- midal. Figure 1. the most prominent of whi ch is w a ter. the Walsh diagram we just developed applies not just to CH 3. the cr(CH 2 ) and the 0 TI(CH 2) d e fin e the C-Hbonding orbitals. depending on w heth er we are considering BH3. orbital A is sli ghtl y s tabilized due to the H • • •H overlap. Toge ther. and so both C' and D a re d oubly occupied. Now orbital Dis occupied in the planar form and D' in the pyramidal form. and the three H's contribute a total of three electron s. in order to take ad vantage of the large stabili zation of C' that occurs. or eight electrons. Moreover. No te th a t sin ce D or D ' is empty in BH3. As in the CH 3 di agram. 1 . now a single TI(CH 2 ). too. a CT u 1(CH 2 ). Let's now use the same procedure for the CH2 group. respectively. resultin g in an important hybrid orbital. The ne t effect of pyramid ali zation is thus des tabilizing. only one p orbital can bond to the hydrogens in the lin- ear form.2 A M OR E MO D E R N T H EO RY O F O RG ANI C BONDI N G 33 Putting the Electrons In-The MH3 System As stated in rule 5 of Table 1. Now. the MOT model is rooted in the most modern theories of bonding. the dominant change is the drop inC. Ou r goals a re the same as th ey were with CH 3 : define the group orbitals and examine how electron popula ti on leads to differing structures. Similar d escriptors for the group orbitals are used as with methyl. Rul e 6 states that we will model the energy as the s um of the one-elec- tron energ ies-th at is. . including a factor of two if th e MO is doub ly occupied. C. we' ll discuss the ramifications of Figure 1. Now.4). water will prefer a bent structure.8 six. we must deposit into Figure 1.3). it has no influence on the geometry. However. the prediction is correct and is identical to the prediction from VSEPR. The only difference is the number of valence electrons. The CT 0 u 1(CH 2) and the p orbital can be u sed to make bonds to o ther groups (see Section 1. B is destabili zed more. These constitute the gro up orbitals for CH 2 .9 applies to all MH 2 molecules. with eight valence electrons.4). the molecule will prefer. We find a cr (CH 2 ) . Here. Putting the Electrons In-The MH2 System The di ag ram given in Fi gure 1. we' ll discu ss the ramifications of Figure 1. so it is non- bonding. and a lone p orbital. Orbital D remains an isolated p orbital. but only a factor of one for singly occupied MOs. CH3. we expect a secondary mixing. consider which geome- try.9 on the geometry of CH 2 as a di screte molecule when we ana lyze reactive intermediates (see Secti on 1.8 on methyl when we look at the geometries of re- acti ve intermedi ates (see Section 1. this time between C and E. converting the planar form to the py ramid al fo rm will be des tab ilizing. and five valence electrons. The six valence electrons of BH 3 occupy MOs A-C of either the planar or pyramidal form. Beca use neutral CH 2 is a carbene. NH3 is predicted to be pyramidal. and so BH3 is predicted to be planar. as it goes from non bonding to bonding. Water ha s eight valence electrons. w here M is a main group element. respectively. BH 3 and NH 3 . and the substantial stabiliza- ti on of D ' associated with pyramidalization outweighs all other considerations. Thus. It is thus the Walsh diagram for BH3 and N H 3. but B and C will be signifi- ca ntly d estabili zed .2. C'. plan ar or p yramid al. 1.4 The CH 2 Group in Detail The m ethyl g roup w as just a start.

then. ei ther bonding model is acceptable. However.10. Th is model is supported by experiment.34 CHAPTER l: I N TRODUCTION TO STRUCTURE AND MOD E L S OF BO Dl C .----. In an isolated water molecule the lone pairs are not identical! Wh at. One is of cr symmetry (C'). and thus has 1T sym- metry. -------. which is the sa me for either model of bonding. and lies significantly lower in energy than the other because it has contributions from a11 oxygen s orbital. In the MO model. the two lone pairs of water (and those of all ethers or other dicoordinate oxygens) are not equivalent sp3 hybrids. as soon as another mole- .--.------~-----~ X W d E' y )-z E E p H H cpt H H C D cr(out) C' ~ B . That is because these properties d e- pend only on the total electron density of the molecule. and they are separated by 2.9 The Walsh diagram for CH2 • The linear form is shown on the left.~ pA Figure 1. In this bond ing scheme the lone pairs are best thought of as MOs C' and D . with two equivalent lone pairs? As shown in Figure 1. of the often invoked model in which water is trea ted as if it has a "tetrahe- dral" structure. and it is converted to the bent form.. Is this model also viable for water? For properties such as dipole moment.2 eV. Our analysis to this stage has been for a single isolated MH2 system. The higher lying lone pair is a pure p orbital. We all know that water has two lone pairs on the oxygen. po- larizability. The computed MOs and the standard group orbital d esignations are also given.- ~ B w A ~ ------. the two equivalent lone pairs can be obtained by taking in-phase and out-of-phase combinations of the C' and D orbitals. Orbitals C' and D could not mix since they have di Hering symmetry. but you may be wondering where they are. etc. However. has some 0-H bonding character. Water in the gas phase has two ionizations corresponding to the two different oxygen lone pairs. the environment that a molecule finds itself in also influences the orbital de- scription. Also shown is a secondary mixing between the C and E orbitals to make C' and E'.

Mixing orbitals is a common tool.. ' . Thjs will allow us to further develop our repertoire of group orbitals. £ ' 0 0. and weakening that overlap raises the orbital energy. it lowers the symmetry of the system. Th e essential features of the two are the same. 1. In summary. Now the C' and D orbitals can mix in MOT and two sp3-like lone pair orbitals are obtained (as in Figure 1. It is now time to describe that procedure in detail.11 Orbital mixing. strong overlap stabilizes an or- bital.. 0.' \ . \ oo .10 ± Mixing the two MOs of water that correspond to th e lone pairs produces two equivalent lone pair hybrid orbitals. £ !'J. and to begin the analysis of more standard functional groups.. 1. Figure 1.1..11.· T£ A.3 and Sec- tion 1. The rules are depicted pic- torially in Figure 1. though. and Figure 1.7.. : : : ' : i'J.11 A describes the mixing of degenerate orbitals. .11 B shows the mix- ing of two orbitals that start at different energies. 0. As m entioned earlier.( 0 '.. and an out-of-phase. Figure 1. i'J. The former is stabilized (lowered in energy).£ . p cule approaches a water molecule. with QMOT and the Walsh diagrams for CH3 and CH 2 . The second-order...10). . The figure shows two mixings: Figure 1. Simple spherica l orbitals are shown to illustrate the concep t. e' > i'J. . nondegenerate mi xing. 3 ORB I TAL MI X I NG. degenera te mi xing. whereas the latter is destabilized (raised in energy). we made use of simple notions: s orbitals are lower in energy than p orbitals.. bonding combination.3 Orbital Mixing-Building Larger Molecules The essence of orbital mixing is stated in rules 8 and 9 of Table 1.11 includes some further detail.. namely alkenes and carbonyls. B.£' > i'J. You should already be familiar with the idea that mixing two orbitals produces an in-phase. where we schematically mix orbitals of two separate molecules. antibonding combina tion (Figure 1. we mixed atomic orbitals to create delocalized molecular orbitals on the groups. the QMOT rules were devised in part on a mixing protocol.. The first-order.BUILD! G LARGER MOLECULES 35 cr(out) "''''·~ ~~ Figure 1. '.£ ' _) 0 i'J. In the end.. This would happen in bulk water. A.7). ·- 00 B. and it follows a precise protocol.

11 A. Fina lly. involving degenerate orbitals. As such.12. A different outcome arises if we have only two electrons in the sys- tem. w hich is summarized in rule 9.7 are degenerate. Figure 1. and because two electrons are raised in en ergy more than the other two are lowered. and the mixing interaction energy should be small (rule 11). and they will be used throughout this text. The statement that first-order mixings are stronger than second-order is really just an aspect of the energy gap law. or " mixing interac- tions". We w ill see in Chapter 14 that the mixings just described can be treated quantitatively using perturbation theory. This statement is known as the energy gap law. . 1. This will be a very important feature in subsequent orbital mixings.11. The cr(CH3 ) and 1r(CH3 ) or- bitals are primarily C-H bonding. As shown in Figure 1. It is always favorable to mix afilled orbital with an empty orbital or to mix two singly occupied orbitals. Hence. the qualitative descrip- tions of orbital mixing presented here are adequate. In this situation. is called a first-order perturbation. using the orbital mixing rules we have just developed. We want to crea te an orbital mixing diagram that combines the orbitals of two CH3 groups to make ethane. for example. The strategy is to build up larger molecules by combining small fragments w hose MOs we understand. the stronger the mixing interaction. The two electrons can come from one in each starting orbital or two in one starting or- bital with the other empty. If both original orbitals are doubly occupied. because they will be the strongest. mixing produces a polarization of the resulting orbitals. these are sh own as not especially strong mixings in Figure 1. why the orbitals Band C of Figure 1. In reality. The high energy. These two aspects of orbital mixing are universal. as this is the geometry appropriate to ethane. and the mixing is stabilizing. since first-order systems have the smallest possible energy gap: zero. the smaller the mixing interaction. rule 11 sta tes that the more the interacting fragmen ts overlap. bonding orbital will have a larger contribution from the original orbital that started out lower in energy.11 B. anti- bonding orbital will have a larger contribution from the higher-lying original orbital. The mixing offilled orbitals on two separate molecules is always destabilizing. we can pro- vide more detailed analyses of some of the issu es addressed here. while the nondegenerate mixing of Figure 1. the overlap in each case sh ould be small. We need consider only the de- generate (first-order) mixings.3. We do need. Now the two electrons end up in the lower orbital. there is a way to gain considerable insight into the bonding of larger molecules. the net interaction is destabilizing.11 B is a second-order perturbation. This difference in energy is true for both mixings in Figure 1.2.36 CHAPTER 1: INTRODUCTION TO STRU C TURE A D MODELS OF BO OJ G A key aspect of orbital mixing is that the antibonding combination is raised in energy more than the bonding combination is lowered in energy. The diagram is set up as in Figure 1.C bonding region.1 Using Group Orbitals to Make Ethane In Section 1. Note that the terms "extent of mixing".7.11 B illustrates an additional feature of the nondegenerate mixing. and it is given as rule 10 in Table 1. for most situations encountered in organic chemistry. Thus. It is generally true that the first-order. It is not at all obvious what the orbitals of methylcyclohexane or acrolein should look like. Also. they do not point out into the C. overlap and energy gap both determine the extent of mixing. the larger the energy gap between the initial p air of orbit- als.3 we built the orbitals of CH3 (actually MH3 ) u sing simple rea soning and symmetry. including. The mixing of Figure 1. We are now in a position to ill ustrate this strategy by combining two methyl fragments to make ethane. when comparing second-order mixings. This four-electron interaction is referred to as closed shell repulsion. This construction gets more difficult to do as the molecules get bigger. however. degenerate mixing is stronger than the second-order mixing. the resulting two orbitals will also be doubly occupied. refer to the energy changes that occur with mixing. the lower energy. We should u se the MOs of pyramidal methyl.12. With the quantitative tools of perturbation theory. one aspect of per- turbation theory that is not evident from the simple analysis given so far. However.

con- sisting of localized C-C and C-H u bonds. All the combinations derived from u(CH3 ) and 7r(CH3 ) orbitals are occupied. 1. and ethane h as six C-H bonds. The computed MOs are s hown in Figure 1. Ethane thus has 14 valence electrons. but slightly C-C antibonding.C bonding region.12 The orbital mixing diagram for the formation of ethane from two p yramidal CH3 groups. and as we place them into Figure 1.7) discussed previously.7. It may at first seem odd that the out-of-phase combinations of the u (CH3 ) and 7r(CH3 ) orbitals are occupied. It is difficult to predict how far this orbital will drop and its relative placement with respect to the other orbitals w ith only the rules of Table 1. CH3 has seven valence electrons.13. Now we add electrons. the u(out) orbitals point out into the C.. The interesting mixing is of the pair of u(out) orbitals. -. Shouldn't these be antibond- ing MOs? They are C-C antibonding. as shown in the figure. but they are still C-H bonding.12.3 O RBITAL M I X I NG-BU I LDI NG LARGER MO L ECULES 37 ~ q3~ + ·. . and this is enough to make them overall bonding. we see that the highest occu- pied molecular orbital (HOMO) of ethane is a degenerate pair of orbitals that can best be described as C-H bonding. The [u(out) + u(out)] MO is the m ajor C-C bonding orbital. the accounting im- plied by the MO treatment is completely consistent with our simpler views of bonding. ·.------------ cr(out) ... They overlap very well. and so the mixing interaction is quite strong. As noted above. The mixing is strong enough that the [u(out) + u(out)] MO drops be- low the out-of-phase combination of the 7r(CH3) orbitals. Again. There are six such orbitals with 12 electrons. Because of the hybridization that resulted from rule 7 (Table 1. so this result must just be accepted at this stage of the discu ssion.. and it has two electrons in it. Figure 1.

h owever. whenever we see a C.12 and 1. transfers consisten tly to all alkane fragments in organic molecules. and it contains a C-C d ouble bond comprised of a a bond and a 'TT bond . .13. we d on' t hybridize. For comparison. and we don' t presume bonding arrangements. In MO T. instead. Moreover. we build ethylene w ith n o assumption s by just combining two CH 2 groups. we can anticipate orbitals of 'TT-like character for the C-H b onding interactions. but now we can see they ma tch up quite nicely w ith the orbitals w e have derived .3.C single bond in a molecule. Prior to our analysis these might h ave seemed to be fairly strange orbitals. Thus.38 CHAPTER 1 : IN TRODU C TIO N TO STRU C TUR E A N D M O DEL S OF BO N DI G cr(out) + cr (out) Figure1. we show the calculated MOs of ethane in Figure 1. we can antic- ipate an orbital of this type.13. This highlights the power of group orbitals. Likewise. [a(out) + a(out)] in Figures 1. their combinations nicely ration alize the origins of orbitals derived from sophisti- ca ted calculations. Let's see how well this construction turns out. 1. experien ce h as shown that the C-C a bonding orbi tal of eth- ane.13 The computed MOs of ethane.2 Using Group Orbitals to Make Ethylene The standard bonding picture for ethylene is viewed as being made from two sp 2 hybrid- ized carbons.

The p orbital d oes not point across the C. ---'' p -p p ci a(oot) ~ a(CH. so ethylene has 12.14 Orbital mixi ng di agram for the formation of ethylene by mixing two CH 2 groups.14. H . The com- bination of these interactions produces the C-C double bond of ethylene. pointing along the C-C bond. we see that the HOMO is the 7T orbital. The crea tion of ethylene from two CH 2 groups is shown in Figure 1.C bond. '.' / ~H1\. ) Figure 1.~--.15.15. the MOs derived from a(CH 2 ) and 7r(CH 2) make four MOs that are primarily C-H bonding. The (p + p) mixing produ ces the 7T bond of ethylene. We ex- pect a strong interaction. When we place 12 electrons into Figure 1.3 ORBITAL M I XI NG -BUILDI NG LARGER MO LE CU L ES 39 H H :u __. The interesting interactions involve a( out) and p. because of poorer overlap. The computed MOs are shown in Figure 1. Hopefully by now the origins of all the MOs are clear. [a(out) + a( out)] is the major a bond component of the double bond. The lowest unoccupied molecular . just as we need for ethylen e.14. A CH 2 group h as six valence electrons. Thus. As such. The hybrid a( out) orbital is strongly directional. the (p + p) interaction should be weaker than [a( out) + a( out)]. 1 . As with eth- ane. The actual MOs are sh own in Figure 1.

LUMO 1.40 C HAPTER 1: I NTRODUCT I ON TO STRUCTURE AND MODELS OF BON DING orbital (LUMO) is the out-of-phase combination of p orbitals. rule 5 states that similar molecules have similar MO dia- grams. but with some changes (more properly termed "perturbations") intro- duced by the oxygen. p.3. Table 1.16. Thus. so the energy of the orbital is lower. the mixings are no longer degenerate as they were for ethylene. these are not experimentally different MOs. P+P A convenient guideline for orbital energies to be used in mixing diagrams is the set of valence state ionization energies given in Table 1. For the particular case at hand. The standard model of bonding in ethylene has been reproduced. The key point is that electronegative elements have rela- tively low-lying atomic orbitals.0 2p 18.2 p 3s 18. and the resulting MOs are given in Figure 1. we can expect similar MOs for formalde- hyde and ethylene.1 Si 3s 17. including both the a and 1T orbitals of the double bond. The precise energies w here we place the oxygen p orbitals are not crucial.8 Valence State Ionization Energies* Orbital -Hii (eV) H 1s 13. Orbitals that are analogous to all the key orbitals of ethylene are created in Figure 1. but there are important differences that we must consider. as stated in rule 12. making it an antibonding 1r* orbital. Note the molecul e is shown in 3p 15 different orienta ti ons for *Technically.4 0 2s 32. The larger the ionization energy.6 c 2s 21. As a result.15 Cl 3s 30 The computed MOs of e thy lene.3 The Effects of Heteroatoms-Formaldehyde p -p In our rules for orbital mixing. However. This is essentially true.8.3 2p 14. it is clear from Table 1.8 that the atomic orbitals that 0 contributes to formaldehyde start at much lower en- ergies than those of Cor H .0 2p 13. 834). We can see the trends expected from our earlier discussions of electronegati vity. as described in Table 1. Formaldehyde and ethylene are isoelectronic.7. they have the same number of valence elec- HOMO trons and the same types of valence orbitals. Although mixing rule 9 now predicts polarizations in all a(out) + a(out) the orbitals.4 2p 11. let's first focus on the 1T and 1r* orbitals.4 N 2s 26. values) used in a popular computational method termed extended Hucke! theory (see Chapter 14.16. Experience h as shown that the primary consequence of introducing heteroatoms into a hydrocarbon system is to alter orbital energies. The mixing diagram for formaldehyde is shown in Figure 1.8 F 2s 40. we must place the oxygen p orbitals below the CH2 p orbital. determined numbers.17.6 3p 14 s 3s 20 3p 11 Figure 1. the harder it is to remove the electron. but rathe r they a re parameters (H.3 3p 9. and we must consider the consequences of the second-order perturbation rules for orbital mixing. . We simply follow the rules of orbital mixing.

o(oot) ? .3 O RB I TAL MI X I NG. Since the building blocks for 1r and 1r* are isolated p orbitals. The lower energy MO. the higher energy MO. the expected polarization may not be evident. you must remember that the atomic orbitals of electronegative ele- ments are sm aller. The actual MOs show that this is indeed the case. So. both the covalent and the van der Waals radii of oxygen are sm aller than those of carbon (see above).17. there is also no ambiguity about the expected polarizations. As such. m ust be polarized toward the oxygen . 1.17.B U I L DI NG LA RGER M OLE CU LE S 41 : X y )-. When you ex a mine the 1r M 0 of formaldeh yde in Figure 1. The fi nal MOs are given in Figure 1.16 The o rbital mixing diagra m of CH2 = 0 from CH2 plus an oxygen ato m . there is no ambiguity about the rela tive energies of the two initial orbitals: the oxygen orbital lies below the carbon or- bi tal.' Figure 1. the 1r orbital. However. must be polarized toward carbon . . 1r*. Similarly. The expected polarization toward carbon is very clear in the 1r* MO. the 1r MO of form aldehyd e d oes have a larger contribution from the oxygen (numerical rather than graphical description s of the MO m ake this clear).

our QMOT m odel of a simple carbonyl is completely compatible with more co nventi onal bon d ing models. the resonance interaction shown is not unreason- able. Looking ahead. not carbon . The 'lT* orbital is perpendicular to the plane of the molecule but is slightly tilted toward the CH 2 and 0 groups. but rather a lone pair. The canonical (fully delocalized) MOs of oxygen-containing molecules predict the ex- HOMO istence of two distinct lone pairs. which we describe as '1T(CH 2 ) minus the oxygen Pxor- LUMO bital (see Figure 1.17 - The computed MOs of . It im- plies a partial negative charge on 0.. the one on carbon. as shown below. The situation with formaldehyde is directly analogou s to what we saw previously with the water mole- cule. and a partial positive charge on C. it places a negative ch arge on a highly electronegati ve element. models emphasizing VBT concepts would focus on the greater electro- nega tivity of 0 vs. However. the reactivity patterns of a carbonyl. the HOMO is not 7T. In order to explain the polarization we have just discussed. Nucleophiles will interact with the LUMO preferentially at the larger coefficient. as expected. Note that this 'lT* MO "leans away" from the bonding region. we cou ld invoke a resonance argument. MOT d oes not always lead to a simple correspondence to our classical views. Although it creates charge and disrupts octets. we find that the LUMO is 'lT*. namely attack at C from a "backside" direction. The higher energy one is significantly C-H bonding. Alterna tively. An interesting feature of Figure 1. For exampl e. While a single Lewis struc ture is certainly a good model for a carbonyl. C. Again.16 is that when we insert our 12 valence electrons. The second lone pair lies below the HOMO. as is typical for antibonding orbitals. cr(out) + Pz c 0 c 0 Better representation of the rr• orbital As in other cases discussed so far. and the lower energy one is significantly C-0 bonding. are completely con- sistent with this model. and is a-type. The C = O form is preferred-perhaps the structure is best thought of as a 4:1 or a 10:1 mixture of the two-the exact details are unimportant. we expect protonation on oxygen. o te the :6 :o: molecule is shown in different orientati ons for different MOs.42 H A PT ER 1 : I N TRODUCTIO N TO STRU C TUR E AN D MOD EL S OF BO N D I N G We still have 12 electrons to allocate. You can combine these orbitals to produce orbitals that re- semble the classic sp 2 hybridization for the oxygen. The nature of'lT* is completely consistent w ith what we expect for nucleophilic addition to a carbonyl. Figure 1. This electronegativity difference should polarize the bond. Certainly the two forms would not contribute equall y to the struc- ture of a carbonyl. protonation at 0 and nucleophilic addi tion to C. the polariza tion of the HOMO toward oxygen has implications for reacti vity. A_ ~ Reso nance in a carbon yl . formed from mix- ing a(ou t) with Pz· Neither of these orbitals are "pure" lone pair orbitals. it should be clear tha t the MO diagram for this prototype carbonyl has significant implications for predicting and rationalizing reactivity patterns. This" tilting" of the orbital h as ramifications for reactivity that we will return to in Chapter 10.16). even a small contribution from the ionic form has significant implications for carbonyl chemistry. However. Also. 8 formaldehyd is a TI-type lone p air with a signifi cant out-of-phase mixture of the 7r(CH2 ) orbital. lt is better represented in a cartoon as shown below.

the MOs for these essential functional groups of organic chemistry can be understood. we would describe the bonding as follows. but it is convenient to make linear combinations of these two. it will be useful to look at its group orbitals in Appendix 3 as a guideline. are those we developed for formaldehyde (Figures 1. Now we should consider at least four different MOs: TI and 1r* along with the two lone pairs. the group orbitals of a simple carbonyl compound. For the terminal methyl groups. while each CH2 contributes two C-H bonding orbitals.10 (in fact.5 Three More Examples of Building Larger Molecules from Group Orbitals All functional groups have a set of group orbitals associated with them.14 and 1. Then. the use of group orbitals provides very valuable insights. This model of alkane bonding is useful. the MOs of a carboxylic acid (Appen- dix 3) can be developed from an orbital mixing diagram. it really does not produce any new insights compared to the more conventional model emphasizing sp 3 hybridized car- bons and simple localized a bonds. This linear combination creates orbitals similar to the sp 3 hybrids available to bond to two other groups. In these instances the TI(CH2) or TI(CH3 ) group orbitals are quite useful. we use the CH2 a(out) and the p orbitals. a methyl ester can be" prepared" by mixing in a 1r(CH3 ) orbital.3. We will show experimental evidence in Chapter 14 that supports the fact that hybridization does not actually occur in the standard sp3. The terminal methyl groups make C-C bonds with their respective a(out) orbitals. For example. designated as the HOMO and LUMO in Figures 1. A linear combination of group orbitals to give orbitals reminiscent of our standard hybridiza- tion notions will always be possible.9. Propene In most instances alkyl groups are spectators in organic reactivity.3 O RBIT A L MIXING-BU ILDI NG LARGER MOLECULES 43 The group orbitals of an olefin would be those we derived for ethylene in Figure 1. The most important ones will be the TI and 1r* MOs. but admittedly. Similarly. 1. as we did for water in Figure 1. an alde- hyde or ketone. The resulting orbital would be the C-C bonding orbital discussed for ethane.3. The CH2 group is more challenging to consider because it must make two C-Cbonds. These can be considered to be the group orbitals of the analogous functional group in larger molecules. The important point is that the orbitals developed by the QMOT approach predict the delocalized orbitals found by high level calculations. Frankly. CH3 (CH 2). However. combining the TI MOs of formalde- hyde with the appropriate AOs of an oxygen atom. 1. From the ba- sic building blocks we have presented so far.8 and 1. and the simple VBT model based on hybridization is adequate. and spmanner. one a(CH3 ) and two 7r(CH3 ). Let's do such a mix- ture for a few additional cases. Appendix 3 shows the full MOs of a number of representative small molecules. where hybridization is not usually the result. we will now show that in more inter- esting molecules. You don' t have to. if we only ever considered alkanes. CH2 and CH3 groups will contribute orbitals such as in Figures 1. So. To do this.17).16 and 1. the a(CH3 ) and two TI(CH3 ) orbitals would be used for C-H bonding. Once we see how the C-C bonds are made. . When considering the reactivity of a specific functional group. each CH3 group contributes three C-H bonding orbitals. sp 2.14. in a linear alkane. The methyl a(out) orbital would be used to make a C-C bond. or a carbonyl. a carbocation. such as an alkene.CH 3. our discus- sion of orbital mixing in water can be extended to any MH2 ) . important interac- tions between the alkyl group and the TI system may occur. for a CH2 or CH3 that is adjacent to a conventional TI system.15.1£ we consider a linear alkane. we just assume them directly. In addition. we have a backbone of C-C bonding orbitals. one a(CH 2 ) and one 7r(CH2 ). However.4 Making More Complex Alkanes In alkanes. 1. there would be no need for CH2 and CH3 group orbitals. We just mix a(out) and p to give orbitals that point in the correct direction to make CH2 a tetrahedral center.

Note the clear prediction that the olefin 'IT orbital and the TI(CH 3 ) group orbital should be out of phase in the " 'IT" orbital.. and so are lower than a C-C 'IT bond. it is enforced by the close juxtaposition of the TI(CH 3 ) orbital with the 'IT bond due to the presence of the C. the HOMO of propene. and (b) slightly de localized onto the methyl group.-·· u u rr(CH ) + rr mol ecule. The predi cted elevation in orbital energy is fully supported by the fact that the ioniza ti on energy of propene (9. the mixing is con- firm ed by experiments. Indeed.18). a factor not usually of importance in understanding reac- tivity. it is not just a coi nciden ce that two of the C-Hbonding orbit- als for CH3 are called TI(CH3 )..18. producing two new propene MOs. and these can interact with the more conven tional 'IT sys- tems of alkenes. This p rediction is confirmed by the actual MO (Figure 1. the 'IT bond of propene is (a) sligh tly higher in energy.4.C bond be tween the methyl and the alkene of propene. Let's consider the consequences of this mixing. a similar mixture can be done with the 'IT* orbital of the alkene that does represent hyperconjugation. there really is no way for the CH3 to interact w ith the 'IT system of propene. and we want to know if favorable or unfavor- . carbonyls. as in Figure 1. Figure 1. An important lesson of this analysis is that there are orbitals of 'IT symmetry in simple alkyl groups su ch as CH3 and CH 2 . These groups orbitals are of the correct symmetry to interact with conventional 'IT orbitals of the sort seen in alkenes and carbonyls. and so on. The mixing w e have just discussed is not hyperconjugation. This mixture is a standard second-order mixing. In addition.51 eV). We are going to focus only on the 'IT system. meaning that the HOMO of propene is higher in energy. and you are asked to show this in an Exercise at the end of this chapter. However.18 shows an orbital mixing diagram for propene. Compared to ethylene. The TI(CH3 ) group or- bital has been lowered in energy. 3 As an example. so we s tart with the ethylene 'IT MO of Figures 1. In our orbital mixing diagram.18. Note that we have mixed fill ed orbitals in Figure 1.~ Figure 1. However. In any particu lar conformation of propene. along with the computed MOs of the 7u 0. There is a distinction in mixing filled orbitals that must be kept straight.. In conventional views of a methyl group. it will always be true that one of the TI(CH3 ) group orbitals will be able to mix w ith the e thy lene 'IT MO. they are C-H a- bonding.18 An o rbital mixing diagram for propene.44 C HAP TER I I NTRODUCTION TO ST RU C T U R E AN D MODELS OF BOND I NG It 0 . w hich is the mixi ng of C-H and C. the two degenerate TI(CH 3 ) group orbitals are low lying in en ergy.15. However.1).14 and 1.73 eV) is measurably lower than that of ethylene (10. If we are consid- ering a possible reaction between two species.C bonding orbitals with adjacent empty or p artially empty 'IT a nd p orbitals (see Sec- tion 1. this interaction is destabili zing because four electrons go into the resulting orbitals. which may seem like a violation of the previous statement that such mixings are always destabilizing.

We could construct a complete diagram.19. Actual calculated MOs are shown in Figure 1.3.p cr(ouQ .8) allow us to place the Cl orbitals below the carbon orbitals. The molecule was not really formed by mixing fragment orbitals. in that there is a nodal surface that splits the C-Cl bond in two. and is especially accentuated at the carbon end of the molecule. There is considerable orbital density on the "backside" of the carbon.B U IL DING LARGER MOLE C ULES 45 Jo+_ . we clearly want to maximize two- electron mixings and minimize four-electron mixings. The C-Cl bond is formed by the mixing of rr(out) of the CH3 with a chlorinep orbital. The most interesting MO is the LUMO. For a destabilizing mixing to occur within the molecule. we are performing an after-the-fact analysis of the molecule.pis th e LUMO. It is C-Cl anti- bonding. we consider methyl chloride as a prototype of an organic molecule singly bonded to an electronegative element.. thus stabilizing the molecule overall.19. but we can also focu s only on the C-CI bond as the most interesting feature of the system. and the antibonding combination polarized toward C.3 ORBITAL M I X I NG. there must be other a and TI bond stabilizing interactions that more than compensate. cr(out) + p able orbital interactions develop during the reaction.+ Figure 1. c Cl c Cl Better representation of the cr• C-CI orbital . It is the prototype a* orbital. Both electronegativi ty (Table 1.. we are doing something different in our analysis of propene. cr* antibonding orbital. - cr(out).18 for propene.1) and orbital energy arguments (Table 1. However. trying to understand the results of a full quantum mechanical calculation on the system. 1 . This means that the polarizations should have the bonding combination polarized toward Cl. using all the CH3 group orbitals and the AOs of chlorine. cr(out) + pis the bonding orbital.19 Orbital mixing diagram for CH3 Cl. The mixing diagram is shown in Figure 1. with computed orbitals shown. and hence a cartoon of the rr* orbital is better represented as shown below rather than as given for generic rr* orbitals shown in Figure 1. such as that shown in Figure 1. A great deal of or- bital character lies outside the region between the C and the Cl. cr(out) . '# . Methyl Chloride As another example of the use of orbital mixing strategies for understanding structure and reactivity. Really. we have simply found that such analyses lead to the true MOs.

as shown in Appendix 3. so that the p orbitals that make up the molecular orbitals appear only as spheres. and so it is worthwhile to present them here. u systems. Yet. and so is formed from only C1 and C3.21 presents th e MOs of benzene. we need to know the group orbitals of some representative sm allTI systems other than just ethylene. We now give the MOs of three oth er essential building blocks of organic chemistry: benzene. and calculated orbitals are given.~] · ~ 6. and we will have to wait until Chapter 14 to see w hy. the simple mixture gets the nod al properties perfectly correct. [ " . and doubly occupied (HOMO) in allyl anion. The HOMOs of allyl and benzyl are completely consistent with resonance models for the two (shown below) . we again need to linearly combine the appropriate group orbitals.22 shows the group MOs of allyl.20. and they mix in the m anner shown to give four mo- lecular orbitals. .·6 * Resonance in allyl and benzyl . and the fact that it is substantially polarized toward the backside of the carbon atom is nicely consis tent with the observed inversion of configuration (backside attack) seen in SN2 reac- tions. and Allyl We have focused our discussion to this point on group orbitals for small alkyl fragments. To a good approximation.. nucleophiles have high-lying filled orbitals that can mix with low-lying empty orbitals. In eva luat- ing organic structure and reactivity. singly occupi ed in allyl radical.CJ. In the S 2 reaction. To build many complex -rr systems by combining small -rr systems.6. there is activity only at the benzy lic carbon and th e ortho and para positions of the ring. The -rr and -rr* or- bitals of ethylene are our starting point. Once again. and allyl. but also the four group orbitals for any di en e. Here. the singly occupied orbital of benzyl radical. Also shown is the "HOMO" of benzyl-that is. a significant fraction of important functional groups in organic chemistry have complex TI system s. The most dis tinctive feature is that the middle MO ha s a node through the central carbon. In benzyl. Benzyl. As we will see. benzyl. Yet.6 Group Orbitals of Representative -rr Systems: Benzene. We can view this as a group orbital for a C-CI functional group that can be considered present in any alkyl chloride. and a few sirn ple TI systems. the empty orbital of benzyl cation. or •) only on Cl and C3. the MO has the sa me form for all three structures. and the dou- bly occupied HOMO of benzyl anion. we show the orbitals from the top. Allyl resonance places the charge(*=+. 1. Figure 1. Note that the simple schematic mixing of the eth- ylen e -rr and -rr* orbitals does not predict the relative contributions of each p orbital to a given MO. a nucleophile will react with CH3 Cl. Figure 1. both in symbolic form and as produced by an accurate quantum m echanical calculation. these units show up repeatedly. wh ere the orbital h as finite coefficients.46 CHAPTE R 1: I TRODUCT I ON TO STRUCTURE AND MODELS OF BO DI NG This u* orbital leads to predictions about reactivity. Butadiene If we are going to build up larger -rr systems than are present in simple alkenes and car- bonyls. This is the MO th at is emp ty in allyl cation. we emphasize that the MOT ap- proach does not require the "patch" fix of resonance to ge t the TI properti es correct. This is done for th e -rr system of butadien e in Figure 1.3. These resulting orbitals can be considered to be the -rr MOs of butadiene. The u* orbital is the appropriate empty orbital on CH 3Cl.

3 ORBITAL M I X I NG-BU ILD I NG L ARGER MOLECULES 47 '' ' .20 The mi xing di agram for creating the 1T MOs of butadiene from ethylene group orbitals and th e calculated orbita ls for the 1T system of butadiene. 1. . '' ' '* ' ' ' ' ' ' ' Figure 1.

21 A. The computed MOs of benzene and conventional representa tions. B.48 C H APTER 1 : I NT RODUCT I ON TO STRUCTURE AND MODE LS OF BON D I NG A. . Two vie ws of the key MO of a benzyl gro up and the conventional representation. B. = Empty in benzyl cation Half-filled in benzyl radical Filled in benzyl anion Figure1.

and amide functional ~ • ..1t can mix onl y w ith the in-phase combination of Cl and C3. These fun ctional groups have 'IT systems that can be viewed as isoelectronic with allyl anion. The resonance structures shown in the m argin highlight this concept. The new mixing produces two new MOs.23 An orb ital mixing approach to forming the MOs of an allyl fragm ent. and such an analysis is shown in Figure 1. 1. these nonadja- cent orbitals d o not overlap significantly. so there is no energy split. ester.7 Understanding Common Functional Groups as Perturbations of Allyl e One of the most useful insights to be derived from th e group orbitals of allyl is a picture of the gro up orbitals for the carboxy lic acid.? Appendix 3 shows the actua l 'IT systems for these functiona l groups.3.. but to a good approximation.BU I LDING LARGER MOLECULES 49 Empty in allyl cation Half-filled in allyl radical Filled in allyl anion -tt- Figure 1. e·v groups. ·· e--" eOH(R) :Q - • • -. yet the general nodal proper- Allyl anion analogs ti es of allyl are retained.... because it lies on the node of the out-of-phase combination. ·· e e '9'0' NH2 in various ways due to the addition of the heteroa toms... We can form the in-phase and out-of-ph ase combinations of p orbitals.22. :- Figure 1. it is simplest to first mix Cl and C3. The result is the allyl pattern of Figure 1.3 ORBITAL M I X I NG. .23. 1. For allyl.. :O~OH(R) • . The three TI MOs of the functional groups are polari zed :O~NH 2 • .22 The MOs of an allyl fragment . we show the p orbitals viewed from the top.. Here. while the Cl / C3 out- of-phase combinati on comes across unchanged. Now we bring in C2. It is an interesting exerci se to form the MOs of allyl using an orbital mixing strategy. where it is clear that the allyl nature is retained. so all we see is a circle.

We already noted that a QMOT treatment of the two structures produces qualitatively similar results. .. is occupied.3. The molecule has 12 va lence electrons. we often need other concepts to fully explain the bonding in a given molecule. A. '''····s .two electron bonds. In organic chemistry. Simple boron compounds are isoelectronic with car- bocations: BH3 has the same number of valence electrons and orbitals as CH3 +. The starting point for understanding the three center-two electron bond is the chemis- try of b oron-containing compounds. B2H 6. resonance is such a concept. ± o- Figure 1. The stability of this well-established bonding pattern can be w1derstood from simple MO arguments. beca use diborane has two three cente r.. The bo- ranes and related structures use the three center-two electron bond extensively. the mixing diagram we developed for allyl translates directly to the H A. Another bonding situation that requires a special treatment for VBT but not with MOT is the three center-two electron bond. forming a pair of three center-two electron bonds. -· c. Borane dimerizes to produce dibora ne. The two MOs of part B a re in-phase a nd out-of-phase combinations of this MO. the three center-two electron bond is asso- ciated with electron-deficient systems. so we describe it briefly h ere.24 A.H-B arrangement in diborane is another three-orbital mixing prob- lem. two apiece. The calculated MOs primarily responsible for the bo nding of the bridging hydrogens in diborane. B..8 The Three Center-Two Electron Bond Starting from the classic picture of bonds between adjacent atoms.24 Three center. to the two B-H-B bridging systems. those for which there are not enough valence elec- trons to make conventional two center-two electron bonds among all the atoms. Only the lowest of the three com binations. B.B bridges.B orbital combina tio ns.50 CHAPTER 1: I NTRODUCTION TO STRUCTURE AND MODEL S O F BO DI NG 1.two electron bonds. The B. Shown a re the in-phase (left) and ou t-of-phase (right) combinations of the fu lly symmetric B.. H H. there a re two bonding orbitals. creating highly bridged structures.H . like allyl. the most common electron-deficient species are carbocations. BH3 in the gas phase dimerizes to diborane. A qua lita tive picture fo r how a s ingle three center- two electron bond is constructed . As explained.As you may recall from introductory chemistry though.H . C. The three center-two electron bond also figures prominently in carboca- tion chemistry. The remaining four valence elec- trons are partitioned. and this molecule has the uncon- ventional structure shown in Figure 1. Since the re are two B._. As the name implies. required to explain the bonding in molecules with de localized 1T systems. In fact. the fully in-phase MO. so BH3 has the same planar structure as CH3 +. eight of which are used to make the four conventional B-H bonds.

All organic functional groups can be envisioned as consisting of a set of d elocalized orbitals that are combined to build up the whole molecule.two electron bonds rather than a single three center. showing their phase relationships. boranes are electron deficient. partially localized MOs-group orbitals. which we have already seen when discussing CH2 and CH3. Transferable. While a detailed d escription of the bonding in some of the more elaborate systems may seem complicated. and as described in this chapter. and in the case of the non d egenerate second-order mixin g. amides. There was initially considerable resistance to this notion.23. These "group orbitals" will be very u seful in fu- ture analyses. they nicely predict the structures of stable molecules and reactive intermediates. contribute a set of transferable orbitals to a mole- cule. The new concepts that were introduced in devel- oping this second model of bonding are summarized as follows: QMOT. and importantly. ketones. Orbital mixing. etc. However. the open and filled circles were meant to be the "tops" of p orbitals. and so we should not be surprised that the MO patterns in a a system such as we are developing h ere are topologically identical to those of an analogous 1T system. a three center-two electron bond.two electron bond. Given the isoelectronic relationship to CH 3 + . con ventional. a simple olefin always has 1T and 1r* orbitals. For example. highly-bridged structures. two center. is that the distinction between a and 1T systems is not as clear-cut as introductory texts would suggest. How - ever.9 Summary of the Concepts Involved in Our Second Model of Bonding The more modern theory of bonding ex tends the notion of bonding and an tibonding or- bitals that describe localized a and 1T bonds to increasingly de localized orbitals. Note that it is generally better to make two. these include 1r(CH3 ) and 1T(CH2 ) group orbitals tha t can interact s trongly with conventional1T systems. and indeed it does (see the next section). A recurring theme of QMOT. there just are not enough valence electrons to make the requ isite number of con- ventional bonds.BU IL D I NG LARGER M OLE CU LES 51 three center-two electron bond. at its core is the special stability of the three cen - ter-two electron bond. the tetravalent carbon was sacrosanct. aromatics. often referred to as non-classical carbocations (see Chapters 2. The system does the best it can by forming a pair of three center-two elec- tron bonds. predictable orbital polarizations result. they could just as easily be s orbitals or sp 3 hybrids.3 O RBJTAL M I XING . the optimal arrange- m ent involves only two electrons. There is absolutely no doubt that three center-two electron bonding is an important component of the electronic structures of a wide range of boron-containing compounds. or between group or- bitals on adjacent groups. it is now clear that three center-two electron bonds d o contribute to m an y carbocation structures. it should come as no surprise that similar bonding should arise in carbocations. This mixing can be between atomic orbitals on adjacent atoms.. 1. Although formally a bonding is involved.23. .3. In Figure 1. the mixing for the B-H-B system (Figure 1. MOs associated with particular bonds can be formed b y mixing orbit- als from the bonding partners according to the rules of QMOT.24 C) still produces three molecular orbitals in a manner analogous to Figure 1. It is a three-atom system involving only one occupied MO. When such mixing occurs. A set of rules for predicting how orbitals will mix to make d elocalized molec- ular orbitals for m olecular groups or molecules. However. Since only one of the three orbitals is highly stabilized. Functional groups such as olefins. and 14 for discussions). 11. 1 . The three center-two electron bond allows a w ide range of exotic looking. Even CH3 and CH2 groups contribute partially delocalized m olecular orbitals to molecules. the destabilization of the higher orbital is greater than the stabilization of the lower orbital.

However. or has less than an octet of elec- trons.2 and 1. or is highly strained (see Chapter 2). A reactive intermediate in organic chemistry is typically some form of carbon that either does not have the requisite four bonds. radicals. This is the first of two preliminary descriptions of reactive intermediates.4 Bonding and Structures of Reactive Intermediates In this chapter we have focused on the molecular and electronic structures of organic mole- cules and on the theories of bonding that explain them. A major goal of physical organi c chemistry.3. First are pentavalent species of the general formula R5C+. The structures of stable organic molecules are mostly straightforward. Many aspects of carbocation chemistry have been at times quite controversial. and more importantly. we describe the bonding models for basic carbocation structures. h owever. and meth yl anion) are all the sam e to a good approximation . carbocations present a much greater variety and more significant conceptual challenges than carbanions. This further development in- cludes the QMOT model given in Sections 1. carbocations.1 Carbocations From the standpoint of molecular and electronic structure. We w ill see that this distinction is enough to yield very different properties for the three types of reactive interm ed iates. interesting issues arise in describing reactive intermediates that will allow us to further de- velop many of the models described thus far in this chapter. Wh ile not common in solution phase chemistry. We begin our analysis of reactive intermediates here by showing how the Walsh diagrams we have already developed provide a solid foundation for understanding the structural fea- tures of reactive intermedia tes. The nomenclature will become more clear as we present examples below. H ere we will consider the extent to which the simple Walsh diagrams can help us anticipate the fundamental fea - tures of reactive intermediates that will allow us to understand their reacti vities.52 C HAP TER 1: I N TRODU CTION TO ST RUCTURE AND MODELS OF BON DI N G 1. and a key concept in reaching that goal is there- active intermediate. Also referred to as carbonium ions are carbocations that have an important contribu- tion from three center-two electron bonding. A basic assumption of the group orbital concept is that the essential features of an orbital di- agram w ill not change as we alter atom types or the number of electrons.8. Heteroatom sub- stitution is allowed. We generated the group orbitals of CH 2 and CH3 so we could use them as fragm ents to build up and understand larger molecules. but also other concepts such as elec- tronegativity and resonance. in the case of de localized systems. . Our notions of bonding should be applicable to both stable structures and reacti ve structures. Heteroatom analogs of these structures are also important. however. and carbenes are also lacking an octet. the basic orbital diagrams for the prototype structures (methyl ca tion. A carbocation is a s tructure wi th a positi ve charge that is associated primarily with a carbon center or. First. However. Rather than disperse our discussion of the structures of these prototype reactive intermediates throughout the text. we must explain some nomencla ture. 1. carbanions do not have four bonds. Carbonium ions encompass two related types of structures. The primary difference is just the number of electrons we put into the diagram of Figure 1. A key issue for these structures is whether the planar or pyramidal (linear or bent) form is preferred. is to explain reactivity. Thus. methyl radical. which emphasizes thermodynamic stabilities. In the following chapter. a collection of carbons. and it will inevitably lead to some positive charge being on the heteroat- oms. There are two types of carbocations-carbenium ions a nd carbonium ions. they are the fundamental structures for a range of reactive intermediates we will see throughout this textbook. This allows us to directly connect the structures to the theore tical models of bonding we have de- veloped and to compare and contrast the properties of the basic classes of reactive interme- diates. we group them here. Here. while radicals.4. we will evaluate quantitatively the features that stabilize or destabilize reactive intermediates. For example. Carbenium ions are trivalent species with a formula of R3C+. or car- benes. such pentavalent carbons are common in the gas phase. each of these is also a well-defined species in its own right.

on the other hand. This very simple analysis predicts the correct geometry. wh ere a neighborin g bond formed from a Hyperconjugation carbon sp3 h ybrid leans toward the carbocation center. are- sult m ad e possible by the development of stable ion m ed ia (see Chapter 2).17 as- sociated with the CH3 . due to the removal of electrons from the bonding 1r(CH 3 ) molecular orbital. and we might expect this to lead to a short- ening of the C. We would also ex- pect a lengthening of the C. this bond " leans" towa rd the em pty orbital. Table 1. Likewise.consis tent with the nomenclature system. there are two orbitals on a methyl group with 1T symmetry. only the term carbon ium ion is used in con- nection with what is more precisely called a carbenium ion.8. The orbital mixing interaction we have just descr ibed is simply the MOT version of what in VBT terms w ould be called hyperconjugation. carbe- nium ions can be thou ght of as proton a ted ca rbenes (:CH2 + H + gives CH3 +). again sh orter than a typical sp3 . Carbenium Ions The prototype carbenium ion is m eth yl cation. Figure 1. Exa mination of the filled or- bital shows a sort of " 1T bond" in ethyl carbenium. in any given con formation of the e thyl cation one of these 1r(CH3 ) orbitals w ill be of the right symmetry to mix w ith this empty p or- bitaL The difference between this situation and the m ixing in propen e is that the 1r(CH3 ) is filled and the p orbital is empty. given the+ ch arge).4 BO DI G AND STRUCTURES OF REA CT IVE I TE RMED I ATES 53 This distinction between the two types of carboca tions was ch ampioned by Olah. and it is always highly favorable to mix a filled orbital w ith an empty orbitaL The orbital mixing diagram for ethyl carbenium is shown in Figure 1. This is completely consis tent w ith the classic view of sp2 h ybridiza tion an d w hat we know about the planar structure of CH3 +. so we use it in this book. Just as w ith propene (Figure 1.25. calculations p redict a bond length of 1. we expect a significan t amount of the positive ch arge on the CH 3 of ethyl cation. The empty orbital (LUMO) of ethyl carbenium is d elocal- ized onto the methyl group. significantly shorter than a typical bond between an sp 3 and an sp2 carbon (1. In the older litera ture. and even just the substitu tion of an alkyl group will start to perturb the ana lysis. CH3 +_It has six valence electron s (one from each of three H 's and three from C. and this too is borne out by calcul ations. reducing the C-C-H angle to 95°. Just as hydronium and ammonium ions represent pro- tonated forms of w ater and ammoni a.425 A. The lowering of the D orbital is inconsequential. because there are no electrons in it. gives a more sophisticated picture. Hyperconjuga tion is often depicted with structures such as that shown in the margin. 1. Since pyramid alization d estabilizes the B/C p air in Figure 1.sp2 bond. Support for this bond shortening comes from the crystal structure of the t-butyl cation [(CH3 hC+]. As such.H bonds of the CH3. and we will see in Ch apter 2 that the order of stabilities for carbenium ions is 3° > 2° > 1o > methyL Wh y should this be so? The reason is an orbital mixing phenomenon. there are a lot of complications to carbenium structures that require m ore complex pictures of bonding than is apparent from this simple analysis.8 m ore than it stabilizes the single A MO. but as we now describe. CH 3 + w ill pre- fer a planar structure. Indeed. and as we h ave emphasized in this chapter. with the P=orbital remaining em pty. Alkyl substitution stabilizes carboca tions. Another inter- esting effect is the distortion of the m ethyl C-H bond that aligns with the empty orbitaL As shown in Figure 1.18).51 A. The preference for planarity is therefore substantiaL Calculations su ggest th at a CH3 + that is pyramidalized to have tetrahedral angles is less stable than the planar form by about 30 kcal I m ol. These six electrons will fill the lowest three MOs of Figure 1. The essen ce of hypercon ju gation can also be viewed w ith the .442 Along. and high-level calculation s su ggest a ch arge of roughly + 0. We feel the more modern no- m enclahlre has value. and is consistent with other nomenclature. We w ill see exp erimental evidence in support of this in Chapter 2. carbonium ion s can be considered to contain proton- ated form s of methane and other typical. The difference between m ethyl and ethyl cations is the CH 3 group.25 A.25 A . tetravalent carbon compound s. The C-C bonds in thjs ca tion are 1. Also.C bond. along w ith the resulting HOMO and LUMO. there sh ould be geom etry changes.4).

H~"'j ~H H~

Related Interests

''H (Eq . As expected.A. The bonding situation is directly analogous to that of diborane discussed previously. As already mentioned. B. and it successfully recapitulates all the major features of the MOT analysis. they are notreallyvery different from the corresponding orbitals of the" conventional" hyperconjugated ethyl cation in Figure 1.·. The structure is shown..34 A for a typical olefin. along with th e HOMO and the LUMO. The HOMO and LUMO of the cation look very much like the HOMO and LUMO of ethylene.25 C2 H 5 +.425 A . 1 . carbenium ions are especially prone to rearrangement.25 B.3. The scrambling of h ydrogens in ethyl cation suggests some H''~''H role for the bridged structure shown in the margin.37 A H HOMO LUMO ~ . ~ 1. To a good approximation. At the same time. The C-H-C unit involves a three center-two electron bond. several different but similar structures can equilibrate rapidly via such rearrange- . Interplay with Carbonium Ions The structure we have discussed for the ethyl cation is appealing and makes sense. to produce th e HOMO and LUMO of the ion.. B. see Table 1... the C-C bond has elongated (1. 1. 1.~ 1. Shown are a structure for the ion and an orbital mi xing diagram between an empty p orbital and a 1r(CH3 ) orbital. The bridged Bridged ethyl cation species is now considered a carbonium ion.31-1.25 A. In addition. it is simply ethyl- ene that has been protonated directly in the middle of the 7T bond.111H Hl''f ~H H No-bond resonance no-bond resonance form shown in the margin. The bridged ethyl carbonium ion.. The issue is whether this symm etrical. it is well-established that there is a very rapid scrambling of the five hydrogens of the ethyl cation both in solution and in the gas phase (we will discuss these H :(!)-. Often. because it ha s a hypervalent hydrogen. 1.8 Figure 1. H H \_§~·''''H H. and the bridg- ing C-H bonds are quite long. as shown in Eq.54 CHAPTER 1: INTRODU CTION TO STRUCTURE AND MODELS OF BONDING A. and a calculated structure is shown in Figure 1. rearrangements in Chapter 11). <~_/.3) H H The structure of the bridged ion (be it transition state or intermediate) is intriguing. H H bridged structure should be considered as a transition state or an intermediate. The ethyl carbenium ion. Our discussion of ethyl cation illustrates many issues that are universal in carbocation chemistry.37 A vs.4).

the t-butyl cation [(CH3 hC+] is a classical hyper- conjugated carbenium ion. and Oka. involving several related structures that are similar in energy and that interconvert via processes with very small barriers. ional process.. the prototype being the tri- gonal bipyramidal PH5 . This small ion may be a quantum 284. The H 2 fragment has a slightly elongated H-H bond distance of 0. Stay tuned. " CH5 • : The Infrared Spectrum.two electron system. If we look at the structures and MOs of the two forms of ethyl cation shown in Figure 1. and Parrinello. in other cases the balance may be tilted in the other direction to favor the carbenium ion.. The its dissimilarity to spectra of other simple molecules like potential energy surface for this carbocation is extremely CH 3 +or CH4 seems consistent with the non-standard flat. However. 375. MH 5 compounds are well known. Allyl and benzyl cations are the prototype delocalized carbenium ions (look at Figures 1. For example. some very high-level calculations have led to description of CH5 +. 216 (1995)." readily generated in a mass spectrometer. However. In fact. such as solvent.21 Band 1. but as the next Going Deeper highlight describes. and to keep the big picture in mind. ture" is poorly defined. The CH5 +ion undergoes a rapid scrambling gas phase infrared (IR) spectroscopy study reported in process in which all five hydrogens become equivalent. No hypervalent atoms are involved in any of the resonance structures for allyl or benzyl. It is planar at the cationic carbon with 120° bond angles. This is a challenging situation. "Structural Quantum Effects mechanical species. White.The complexity of the spectrum and kcal / mol.746 A for the isolated molecule). T.25. . 0. The molecular structure given for CH 5 + is reasonable for discussion purposes. The same orbitals that carry the negative charge in allyl and benzy l anion carry the positive charge in allyl and benzyl cation. CH5 + does not adopt such a highly symmetric geometry.However. counterion. The positive charge is distributed around the entire structure. D. Many of the most contentious debates in physical organic chemistry have boiled down to efforts to sort out very subtle structural and energetic differences on a given carbo- cation potential energy surface. In fact. Tang. This rapid equilibration is a manifestation of the fact that the potential energy sur- faces of carbocations are often very flat. Carbonium Ions The prototype carbonium ion. Conventional1r delocalization does not convert a carbenium ion to a carbo- nium ion in this case. it does not represent the true "nature of the beast" . M. The easiest way to think about this struc- ture is as a pyramidal CH3 + bonded to a molecule of H 2 . Going Deeper CH 5+-Not Really a Well-Defined Structure rigid geometrical object is simpl y not applicable. 135 (1999). we might expect that such situations would be s trongly influenced by context effects." Na ture. exact substitution pattern. Marx. a very sophisticated complicated. However. full assignment of the the conclusion that there is effectively no barrier to the flux. and so on.. with very strong acids like FS03H. A representative structure is shown in the margin. protonated alkanes can be generated. spectroscopic characterization of such a "representative" because the situation is actually fairly struch1re is challenging. CH5 +. Our goal at this point is simply to alert you to the potential for such complications. Not We have just described the structure given for CH 5 +as surprisingly. J. for which the classical model of a and 3-Center 2-Eiectron Bonding in CH 5 • . makingCH 5 + a highly flu xional molecule. This makes a three center-two elec- tron bond of the sort we have described for BH3 .22).4 BO N DI NG AND ST RU C T U RE S OF REA C TIVE I N T E RM EDI ATE S 55 ments. making CH5 + an important model compound. temperature. with several possible isomers that differ very little in structure and energy. 1.87 A (vs. in which case the notion of" molecular struc. T. the differences are certainly small. 1999 found a complex collection of lines that have been Estimates are that the barrier to this process is less than 1 assigned to CH5 +. The CH3 + contributes an empty CT out orbital that interacts with the filled CT bonding orbital of H 21 making the three center. while the C-H bonds involving these hydrogens are substantially elongated. many calculations indicate that the classical non-bridged structure is best thought of as a transition state for the hydrogen scrambling reaction. In the specific case of ethyl cation. In addition. high-level theory actually favors the bridged struc- ture as the lowest in energy by severa 1kcal / mol. Also. E. lines was not possib le. carbocation potential energy surfaces are often flat.

Figure 1. and 14. The b ridged picture of ethyl cation has a hypervalent hydrogen. Two classic examples are cyclopropylcarbinyl cation and nor- bornyl cation. As a result. and so. Some of these com- pounds are called non-classical cations.with a to tal of eight valence elec- trons. This stabilizes more electrons in the pyramidal structure of an anion because the 0 o rbitals are populated. This is more difficult when a carbon is sp 2 hybridized (favoring 120° a ngles) than when the carbon is sp3 hybridized (favoring 109S angles). Many other ~H carbocations are also bridged species best defined as carbonium ions. An alternative ra tionaliza tion can be developed by referring to the CH3 Walsh di agram. Returning to the Walsh di agram (Fi gure 1.2 kcal I mol for simple systems. and thus raises the inversion barrier. the lone p air is in a p orbital. 1. roughly 1. 3 R u.4) R2 R... The prototype structure of a simple.~ . as is observed experimentally. This effect raises the energy of the planar transition state for inversion in the small ring relative to an open system . again. Thus. This electronegati vity effect is general.8 pred icts that H 3 should be py- ramidal and BH3 should be planar.3). the lone pair occupies a hybridized orbital and the carbanion center Cyclopropyl anion would be roughly sp3 hybridized. As noted previously. This high barrier can be understood by recalling that electronegative substituents prefer to bond to orbitals with grea ter amounts ofp character. a basic tenet of the group orbital concept is that the MOs ofMH 3 are essentially the same. electronegative atoms low er the energies of all the MOs to which they contribute. because p orbitals in general are easier to w ithdraw electrons from than s orbitals. We will discuss these structures much more in Chapters 2.. NF3 has a barrier of ~ so kca l I mol. the C-C bonds must be bent to accommodate the 60°bond angles of the cy- clopropane ring. e ~ (Eq. which h as no contribution from attached fluorines. Several factors can significa ntl y raise the inversion barrier at a carbanion cen ter. 1. 1. raising the inversion barrier. on m ost reaction time scales. Although ethyl cation fits this description. Consider the inversion of the cyclopropyl anion. the term is normally associated with s tructures where a C-C cr bond brid ges to H A H H "fi~ H Norbornyl cation H create a hypervalent carbon.3. In the py ra- midal ground state. The 0 orbital of Figure 1. One is incorporation into a small ring.4). This stabilization is enhanced by further mixin g to pro- duce 0 ' . cyclopropyl anions often react as if they were stably pyram idal. Thus. 11. simple carbanion s sh ould be py ramidal. However. Electronegative substituents attached to the anionic cen ter can also substantially raise the inversion barrier. again consistent with experimental observations. Whi le ammoni a has an inver- sion barrier of ~5 kcal l mol. shown in the margin. These are defined as containing bridged structures Cyclopropylcarbinyl cation with three center-two electron bonds and a hypervalent atom. CH3 . In both the ground and transition states. . electronegative substituents preferentially stabilize the sp 3 ground state over the sp2 transiti on state. An equivalent way to think about the structure is as an sp3 hy- brid.4. pyramidal carbanion h as a lone pair of electron s in the crou t orbital of Figure 1. carbanions behave as if they were effectively planar (Eq .8. because they all have contributions from the fl u orines.56 CHAPT E R 1: INTRODU CT ION TO STR UCTU R E AN D MOD ELS OF BON DI NG As discussed. We have already noted the isoelectronic relati onships between BH3 and CH3 +. regardless of the identity of M. all th e orbitals are lowered in energy. including 0 . and the stabilization it gains by pyram id aliza- tion d ominates the Walsh diagram.8 is now filled. placing the lone pair in an sp3-like orbital. and the carbanion center is essentially sp2 hybri dized. On going from NH3 to NF3 we replace H by the high ly electronega tive F. an R5 C+ species can be considered to be a brid ging structure w ith a h yper- H valent carbon. and between NH3 and CH3. In most cases the barrier to inversion at a carbanion center is small. the effect of fluorine on the planar structure is to lower all the orbital energies except D. w ith the pyramid a l s tructure. Thus.8).2 Carbanions Now let's consider the proto typical carbanion. As noted in th e di s- cussion of formaldehyde (Section 1. the only difference being the number of valence electrons. In the planar transition state for inversion.

NH3 is p yramidal. and it is best to say that no obvious preference for planar or pyramidal geometry can be predicted. we would expect the delocalization (resonance) effect to win out. and the stabilization it provides is to some extent offset by the destabilization of other. The net result is that the parent methyl radical.. This preference is because the interaction w ith the 1T substituent will be greatest in this form. the singly occupied molecular orbital (SOMO). where we have a pure p orbital rather than a a(out) orbital to participate in the delocal- ization. PH3 The magnitude of the stabilization increases as the initial The pyramidal inversion b arrier. e- 0: 1. allyl and benzyl radicals are stabilized by delocalization. but 35 kcal / mol. CF3 • is very strongly pyramidal for the same reason as discussed above for NF 3 • Electronegative substituents prefer bonding to an sp3 h ybrid rather than an sp 2. First. Examples of such substituents that will lead to a planar neighboring anionic carbon are cyano. Hc e~o-. a ouu is only singly occupied (Figure 1. is essentially the same as the MO that is doubly occupied for the respective anions (see Figures 1. and carbonyl. doubly occupied orbitals. is a cru cial property energy gap be tween D and E decreases. and we expect significant negative charge only on those atoms that have significant contribution in the HOMO.4 BO N DI NG AN D S TR U CT U R ES OF R E A C TIVE INTER MEDI AT ES 57 Connections Pyramidal Inversion: NH3 vs. but the barrier is so low n egative than N. CH3 •. Th e geometries of such species can also be understood using our notions of bonding.22). One example is the one- electron oxidation of an alkene. However. E. This greatly stabilizes weakly. It is difficult to predict the geometry of this species based on the Walsh diagram. is planar. The crucial molecular orbital.8). NH3 and further diminishing the initial D/E gap.. Two factors make of any pyramidal tricoordinate molecule. Pis less electro- previously. As discussed this gap smaller for PH3 than for NH3 . so phosphines are stable pyramids at con. as simple radicals show only a very weak preferen ce for the planar structure. Given the small intrinsic preference for a pyramidal structure. diminishing the gap. the mixing of the alkyl group's TI-like group orbitals with the now singly occupied 1T orbital becomes even more pro- . There are also reactive intermediates known as radical cations. Resonance structures make this argument clear. Second.21 B and 1. H2 C ~ . oxi- dation retains a planar structure in the alkene. and simple substitution can produce pyramidal radicals. In fact. Stabilized carban ions midalization. the 2s and 2p orbitals in ture. nitro.8). however. at distances To rationalize this difference between NH3 and where the phosphorus 3p orbitals interact strongly w ith PH 3. where the electron is removed from the 1T orbital. Substituents that stabilize a carbanion by 1T delocalization will favor the planar s truc- ture. v. Allyl and benzyl anions are also planar. PH3. Just as with allyl and benzyl cations and anions. is" consis- tent" with experiment. making the anionic carbon planar. The HOMOs of these anions are shown in Fig- ures 1. respectively.4. thus (~ s kcal / mol) that amines act planar at room tempera. The crucial orbital that is stabilized by pyra. the P 3s orbitals inte ract relati vely to generate D ' and E' (Figure 1. 1.21 Band 1.. and thi s is discussed in Chapter 2.. so D will be higher-lying in PH3.22. making E less anti bonding in PH3 than in D ' and is a key factor favoring the pyramidal form. and this preference completely shifts the balance toward pyramidalization in the case of the radical system. The first is an electronic effect of the sort discussed previ- ously for anions.'"' on the order of the first row of the periodic table have similar sizes. we focus on the key interaction between D and E the H l s orbitals. beyond the first row the s orbitals are significantly more ventional temperatures. The second reason that radicals are pyramidal is a conformational effect. but the energy cost for distorting away from planar- ity is sm all. Two factors fa- vor pyramidalization in radicals. Consequently.3 Radicals Methyl radical has seven valence electrons. too. This. extensive study has revealed that the intrinsic preference for planarity in methyl radical is so small that essentially all other localized radicals are not planar. h as a very high E. contracted than the p orbitals. - 2 . For all al- kenes besides ethylene (see Chapter 2 for a discussion of the radical cation of ethylene).

As the H-C-H angle contracts. and radicals-the Walsh diagram for CH3 provides the starting point for the discus- sion. let's consider carbenes.9. With small bend- ing. Experimentally it turns out that the triplet is the global ground state in methylene. However. a gap opens up between C and D. For the reactive intermediates we have considered so far-cations. some predictions are still possible. While simple carbenes have a triplet ground state. The Walsh diagram of Figure 1. the SOMO is even more delocal- ized onto the neighboring R group. Carbenes with lone-pair donating substituents such as N. appropriate substituents can reverse this preference. We might expect its reactivity to be similar to that of radicals. and we can place two electrons each in the A and B MOs. Its reactivity patterns should be quite different from radicals. It also contains an empty p orbital. The most effective way is to interact with the empty D orbital of the singlet. the triplet state should have a wider H-C-H angle than the singlet. there is no large benefit to pairing two electrons into C. the other reactive intermediates we have studied. :CH2 . what we have called panda( out). Each car- bene is in fact two reactive intermediates. the angle is 136° for the triplet and 105° for the singlet. and indeed this is the case.9 provides an excellent starting point for considering these issues. as shown in the margin for the propene radical cation. differentiated by the spin state of the system. If the spins are aligned. as shown in the margin. anions. This is a triplet state. This is a singlet state. and the triplet is sti ll preferred. This leads to a unique feature of carbenes vs. then the spin state of the system isS = ~ + ~ = 1. indifluorocarbene the singlet lies below the tripl et by . Hund's rule predicts the high-spin. These two states are expected to have substantially different molec- ular and electronic structures and to show distinct reactivities. . The effect can be quite large. The Walsh diagram provides a satisfying analysis of the electronic structure of carbenes.58 CHAPTER 1: INTRODU CTION TO STRUCTURE AN D MODELS O F BOND ING nounced than we saw for propene (Figure 1. 0. by approximately 9 kcal/ mol. The bonding and structural model we have developed provides excellent guid- ance as to how we might create a carbene that has a singlet ground state.50 kcal I mol! Detailed theoretical studies reveal a linear correlation between the singlet-triplet Resonance in singlet carbenes gap and the electron pair donating ability of the attached substituent(s). and the essential features of the system can be summarized by the simple representations shown in the margin. just Carbene electron configurations like a simple carbenium ion. where a bracket is used to denote one electron and a formal positive charge in the MO. the relevant electronic structure questions can be considered by referring to the Walsh diagram for CH2 in Figure 1. neutral :CR2 species. Therefore. If we keep one electron in each MO. leaving two valence electrons to occupy the degenerate C / D pair. For carbenes. It is impossible to unambiguously predict the absolute ground state of m ethylene wi th simple models such as Walsh diagrams. triplet state should be preferred at the linear geome- try. It contains a lone p air of electrons in an MO [a(out)] that is reminiscent of an sp 2 hybrid. The singlet state is quite different. a resonance model also nicely rationalizes the stabilizing effect of donating substituents on carbenes. such that some substituted carbenes show a large energetic preference for the singlet. As is typical. This is indeed the case. For example. 1. and the multiplicity is m5 = 25 + 1 = 3. However. All simple dialkyl carbenes have triplet ground states.4. We begin with the linear form on the left side of Figure 1. The triplet has two electrons in two very different orbitals. As w e will see in Chapter 14. and we will see that they are (Chapter 10).18). understanding and predicting spin preferences requires more advanced treatments of electronic structure than we are providing here. Methylene has six valence electrons. and indeed it is. and halogens can have singlet ground states because of such an interaction (see the next Going Deeper highlight for an example). the prototype of which is the mole- Radical cation of propene cule methylene. and a singlet with both electrons in C w ill become the ground state. S = ~ + (-~) = 0 and ms = 1. If we have opposing spins. when the angle becomes small enough.9.4 Carbenes Lastly. this distortion should be mildly stabilizing. the lower energy of orbital C will over- come the electron repulsion energy.

Therefore. Chem. Thus. [}-H N0 I base C:> N I R R and ind eed these types of carbenes ha ve a singlet ground state. these should grea tl y stabili ze the singlet state. ch emistry (see Chapter 12). A. as is a hybrid orbital along an axis. it is useful in this chapter to develop a sim- ple bonding model for organometallic and inorganic complexes. Likewise. The second effect is electronic. They h ave recentl y found use as ligands for an led to some remarkable new carbenes.26.1. Samples of i are stable for Ard uengo. years. how . Likewise. the dx'-y' orbital is aligned along the x andy axes. These stable carbenes are not solely theoretical ever. Res. The same kind of loca lized and discrete a and 'TT bonds often associated with organic compounds can be assigned to the bonding in organometallic compounds. j. and is therefore "sigma-like". First. and is therefore considered to have sigma symmetry when viewed down these axes. 913. the dxz and dxy orbitals look like two orthogonal p or- bitals. 32. let's examine the shapes and nodal properties of the d orbitals shown in Figure 1. pl e had R = adamantyl. w e refer to it as having sigma symmetry along the z axis." A ce. 1. Two potent electron donors are attached to the carbenic center. Here we examine a model analogous to the first VBT /MOT model of organic bonding given in Section 1. We will leave an examination of structure in organometallic systems to Chapter 12. Hence. the dz' orbital and the dx'-y' orbital can make sigma bonds that are placed along the z or the x andy axes. and extensive physical characteriza- We generally think of carbenes as extremel y reacti ve spe. and an x-ray structure reveals an N-C-N angle of 102° at A stable carbene the carbene center. In fact. as long as they are protected from air. Many deriva . IlL "Looking for Stable Carbenes. First is the steric bulk of the R group. In recent years. One simple con- cept for visualizing bonding in m etal-containing systems is to make direct analogies to such bonding in organic systems. these molecules can be crystallized. which is just one of the current chemical subdisciplines where the tools of physical or- ganic chemistry are often applied . When citing down the x axis. when forming discrete localized bonds to ligands. clever appl ication of the concepts described here has curiosities. When creating double bonds to ligands. structure. useful new materials. and for the most part they are. aliphatic rin g system. For example. tive intermediates can lead to fundamental insights and bene. and not just look at organic bonding. . both the dyz and dxy look like p orbitals.921 (1999). citing d own they axis.5 A V ERY QUIC K LOOK AT ORGANOME TAL LI C AND I NORGAN I C BONDING 59 Going Deeper Stable Carbenes tives have been made. As just discussed. 1. We essentially use these orbitals just as we use hybrid or a( out) orbitals on carbon. respectively. research into basic reac- Two factors contribute to the stability of this type of car. In this regard. a large.. these orbitals are considered to have 'TT symmetry. in excellent agreemen t with the expec- tation for a singlet carbene. tion has provided detailed insights into carbene electronic cies. The first exam.. Remarkably. notice that the d2 z orbital is directional along the z axis.5 A Very Quick Look at Organometallic and Inorganic Bonding One theme of this textbook is to consistently tie organic chemistry to organometallic chemis- try. and we will examine more complex MOT ideas about bonding in metal-containing systems in Chapter 14. The breakthrough important class of ruthenium-based olefin m etathesis cata- occurred in 1991 when Ardu engo and co-workers at lysts that h ave profoundly influenced synthetic organic DuPont reported the synthesis and isolation of carbene i. R R Later examples included heavily substituted aromatics as I I N N the R group. Now let's examine the analogies that can be drawn between the rest of the d orbitals and p orbitals.

analogo us to rr(out) orbitals or th e standard hybrid orbitals for making a bonds to orga nic groups. A. For example. the dxLyz orbita l looks like a carbon hyb rid . these orbitals on the metals are used exactly as we use p orbitals on carbons.6 for the lin- combination of a TI-symme try d ear combinations of sand p orbitals to create hybrids. Figure 1.27. B. orbitals.28. Looking alo ng the positive x or y axes.27 Lastly. The a and b descriptors on each orbital refer to the line of sight for the : cor responding eyeballs. a hybridiza tion state know as dsp 3 is obtained. the carbo n hybrid along th e z axis a ppears the sa me as a dzz orbital.60 C H A PT E R 1: INTRODUCTIO N TO STRUCTURE AND MODELS OF BOND ING A. if you start with an sp2 h ybridization state.. As shown in Fig- ure 1. The d. This combina- tion is very clear in Figure 1. and mix the remaining p. the bonds to the apical positions are pd hybrid orbitals. we can mix d orbi tals with the s and p orbital with a carbon p orbita l to create a metal-carbon 1T bond. Using the sam e concepts presented in Section 1. . C.1. They a re : referred to as hav ing 1T symmetry. orbital with the d. Analogous to a p orbital along the z axis or the y axis Analogous to a p orbital along the x axis or the y axis Analogous to a p orbital along the x axis or the z axis Figure 1.w d . there is even a h ybrid orbital app roach to modeling the bonding in organometal- Orbita l mixing di agra m for th e li c and inorganic complexes. where an orbital mixing diagram is given for a 1T bond be- tween a m etal and a carbon.26 Schem atic representations of d orbi tals a nd their analogous carbon orbitals. Analogous to a carbon hybrid along the x or y axes c. They are referred to as having rr symmetry. Such h ybridization is appropriate for trigonal bipyramidal complexes. Looking along the positive z axis (designated as an eyeball). as would be present in an organometallic complex that has an M = CR 2 functional group. while the equatorial po- sitions are sp 2 hybrid orbitals. and du orbitals appear asp orbitals.y. orbital. Analogous to a carbon hybrid along the z axis § B. depending upon the line of sight.

so it can be covered at any point during the course to learn more advanced concep ts and quantitative methods. after analyzing struchue. With our current qualitative models we can go a long way in our analysis of topics in organic chemistry. we took the topics of this chapter only to a depth that is routinely used when thinking about or- ganic s tructure and bonding by a non-ex pert in quantitati ve methods. We presented two models for bondin g. Furthermore. We do ex- actly tha t in the next several chapters where the focus is more upon structure. sp hybrid orbitals Figure 1. with hints at structure and reactivity. it should be apprecia ted that to the bes t of th e authors' abilities. which is appropriate for octahedral co mplexes. For almost all bonding geometries in inor- ganic and organometallic chemistry.___. three p's. hybrid orbitals are useful. Hopefully this chapter has refreshed your m emory. However. For many students and professors. It is only a first look into bonding.5 A VERY Q U I CK LOOK AT OR GANOME TALLIC AND INORGAN I C BOND I NG 61 ---. the orbitals a long ap ica l positions are pd h ybrid s. and not until we have to look at some very specialized reactions (pericyclic and photochemical) will we need to de- velop a more so phisticated theory of bonding. because it is sufficient for almost all organic transformations. Summary and Outlook This chapter represents just a beginning. it may be desirable to now go directly to Chapter 14. and has sparked an interest in you to learn how these notions of bonding can be put to use. and dx'-y' orbitals leads to d2 sp 3 h ybridization. and showed that they can be used to understand stable organic structures and reactive intermediates. Structure and reactivity actually take up approximately two-thirds of this book. However. The mixing of an s. The orbita ls a ligned along the equa torial positions remain as sp 2 hybr ids. this chapter may completely suffice as a review of bondin g. . 1. and the d~. pd hybrid orbitals 2 e = Equatorial positions. a classical one and a more modern approach. and this very brief introduc- tion to bonding using m etals will be enough to take u s a long way in understanding stru c- ture and reactivity. however.28 Hybridiza tion for h·i go na l bipy ramidal geometries starts w ith the ce ntral a tom sp 2 hybridized. This is a decision to be made on an indi- vid ual basis. we can take the bonding concepts and loo k at reactivity. 3 sl Hybrid orbitals pd Hybrids Remains a =Apical positions. For other students and professors. w here the concepts introduced herein are discu ssed more quantitatively and modern methods in com- putational electronic struch1re th eory are covered . Our intention is for Chap ter 14 to stand alone.

Show a ny pl ausible reson an ce structures for the molecules give n in Exercises 1 and 5. and HI.H bond an gle is 118°. /""'. c. 6 3. Consider bond dipoles to predi ct whi ch confo rme r of fo rmi c aci d s hould have the hi gher dipole mom e nt.1).78 D a nd 3. 6. Ide nti fy w hat and whe re tha t charge is. Ex pl ain w h y the dipole moments a re so different. 0 A. indi cate this (n ote th a t the molecul e corresponding to le tter A is the sam e in each proble m). Pauling proposed the fo !J owing correlation be tween electro nega tivi ty difference and percent ioni c characte r in a bond : ' 14 ] where XAis the electro negativ ity of A.e-<xA-x"> HCI. as well as heteroa toms and hyd rogen. 0 E. Indi ca te an y pola r- iza tion these bonds may have in yo ur diagram th rough the shapes of the discrete bonding and antibonding orbi tals. The electronegativities o f the . 0 0 H }-o. Ionic ch aracter = 100 X [1 . fo rmic acid (the highe r dipole confo rmer of Exe rcise 10) or fo rma mide? 13. A B. respecti vely. given th at the H-C. Draw a molecule that h as n o dipole mom ent. A.C = N groups are the sa me (Table 1. # 2. 9. Dra w the TI and TI* molecul ar o rbitals for the discre te and loca lized TI bonds in the foll owing structures. ~N D. For the bond polarizati ons in Exercise 2. H H B. rathe r than the d/ topology. B. but has a qu adrupole m oment with the to pology of a d xy o rbital. draw a dipole arrow for the polarized bond s. but it d oes have an octo pole m oment.92 D. Formaldehyde has a fairly large dipole mome nt of 2. }.33 D. identify which compounds ha ve an a tom that has a formal charge.H bonds of cyclopro p an e. lf there are no pl au sible resonance structures. 4. identify any significant bond polari za tio ns by w riting 8+ and 8. The mos t p o lar bond for the elements in Table 1. Which should have a la rge r dipole mome nt. D. 8.C and C. but CO h as a small dipole m oment of0. Calcul a te the percent ionic charac ter in H F. H H I er I N N 0 0 A. Prov ide a simple descripti o n o f this moment. 5. 14. HBr. In the compound s w itho ut form al ch arges. c. What is its pe rcent ioni c characte r? 12. but the dipole mo ments of CH 3C= CH and CH 3 C =N are 0. Indi ca te any p olari zation these bonds may have in your di agra m b y drawing the sh a pes of the discrete bonding and antibonding orbitals in a manne r indi ca tive of this polari zati on. B. State the hybridization of the no n-hydrogen a to ms in the following structures. A or B. s A. in the following structures.1 would be in KF.C= CH and .11 D . . 0 )lo~ c. Carbon te trachloride h as ne ithe r a dipole mome nt nor a qu adrupole moment. A B.a A. 7. Draw the u and u* molecul ar orbitals fo r discre te and loca lized bo nds formed betwee n h ybridized carbo ns a nd hetero- atoms (a heteroatom is any a tom besides C and H ). Di scuss the hybridiza ti on in the C. 0 E. \_ CI c.nea r the a ppropriate atom s. Use resonance and electronegati vity arguments to explain these results.Br D. H 11. 10.62 CHAPTER 1: I NTRODUCTION TO STRUCTURE AND MODE L S OF BOND I NG Exercises 1. Assuming that each atom in the foll owing stru ctures has an octet of electrons.

385 110. Does an argument based upon electronegativiti es also explain the trends in bond angles? Molecule C-H bond length C-X bond length H-C-H bond angle H-C-X bond angle CH 3F 1.0 CH3Br 1. Describe and draw (using cartoons) h ow the three gro up orbitals for a pyramidal methyl group (Figure 1. electron.58 19.8) can be linearly combi ned to give orbita ls simi lar to the individual u bonds for med from sp 3 hybridized carbon atom s and hydrogen 1s orb itals. use orbital mixing to predict all the orbitals for the 1T system of h exatriene. and then use a hybridization / electronegativity argument.14 CH3l 1. EXERCISES 63 15. Sketch the 1T and lone pair gro up orbita ls for the following organometallic group. for the carbon hybrid orbitals used for the C-H bonds of CH3 Cl (H-C-H an gle of 11 OS). 26. Use the appropriate group orbita ls and the QMOT rules in Table 1. 24. Contrast this qu adrupole moment to that in benzene. starting wi th methylene and OH. Rationalize the trend in the C-X bond lengths (angstroms) . or neutron diffraction-can be applied to larger molecules. The H-P-H bond angle in PH 3 is nearly 90°. whereas that for NH 3 is onl y sli ghtly contracted from the idealized 109. Shown below are bond lengths and angles for various meth yl h alides determined by various methods. dimethyl ether. 23. Also rationa lize the H-C-H and H-C-X bond angles using VSEPR.7 to create the molecular orbita ls of protonated formal- deh yde (CH 2 =0H +). Consid er the methyl C-H bonding in your answer.R M =<R 28.12 107. and the C-H bond s of ethylene (H-C-H angle of 117.3°). 16. and the calculated F-C-F angle in CHF 2 is 112°. Highlight the differences and explain how this diagram would rationalize reactivity differences in alkanes and alkyl fluorides with nucleophiles. .09 1. 29.F u bond in CH 3F. 30. Rationali ze the differences between the 1T orbitals of butadiene and acrolein (the MOs are shown in Appendix 3). The calculated H-C-H bond angle in methyl radical is 120°.methyl single bond using the dz2 orbital on the metal and a u(out) orbital on the methyl group. crea ted by the approach of two pyramidal methyl groups.781 110. Sketch the hi ghest occupied MO of propene and compare and contrast this MO with the an alogous MO of ace taldehyde (CH 3CH = O). 27. as they do in benzen e. 20. Geometries for small molecules such as methyl fluoride can be determined from microwave spectroscopy. 22.096 1.SO. where M is a second row element such as Cor 31. Rationali ze the differences between the fo llowing dipole moments for water.) O.52 108. Consider the methy l C-H bonding in your answer. (Hint: Use the group orbitals of formic acid as your guide. Use the same energy axis for both plots. Sketch the 1r* MO of propene. if carbon were signifi can tly more electronegative than hydrogen.2 CH 3 CI 1. the ca lcul ated H-C-F angle in CH 2 F radical is 115°.096 2. Draw mixing diagrams using u(o ut) orbi tals that compare the fo rm ation of the C-C u bond in ethane. and ethy lene oxide (oxi rane): 0 D 1.939 111. .2 108. Use an orbita l mixi ng diagram to rationalize the nature of the HOMO of methylamine shown in Appendix 3 . Convi nce yourse lf that the C-H bonds of cyclohexa ne wo uld add up to give a quadrupole moment. Draw an orbital mixing diagram for a metal.11 1. Create the group orbitals for an MH system. The results of Figure 1. sp".20 should be a useful starti ng point. 18. 21. to the for mation of the C. 25. whereas vari- ous forms of diffraction-x-ray. Explain this difference using a VSEPR argumen t.139 111. Starting with two ally l gro ups. Predict the hybridization state. Rati onalize wh y the structures become more pyramidal with increasing fluorine substitution.5 106. called a Fischer carbene.89 17.

A. Watson. Molecular Structure and Bonding: The Qualitative Molecular Orbital Approach. A Useful Compilation of Standard Bond Lengths Alle n.. W. Soc. and Taylor. Hybridization. will the more electronega tive ligands prefer the apical or equatorial pos iti ons? Explain your answer. Epiotis. Prentice--Hall. 33. Perkin Trans. R. B. 34.. organi c chemistry. W. and Salem. (1973) . Gimarc.). Chapte r 4. . Wheland. and Group Orbitals Albright. 35. Springer-Verlag.. and discuss the diffe rences. Borden. L. Combine the lone pair orbitals on formaldehyde to achieve what resemble the classic sp 2 hybrid lone pairs. Further Reading Much of the material in this chap ter is review. Part 1. H . Bodrowicz. F. S1-S19 (1997). an Introduc- tion to Modem Structural Che111istry. L. (1960) . Bra mmer. Jorgensen. (1983). Modern Molecu lar Orbital Theory for Organic Chemists.. 0. 1977. Methods of Electronic Struc- ture Theory. and if you are not familiar with the material in Section 1. New York. Kennard. Sketch the 'IT orbitals for allyl anion and for an enolate anion." f.. M. Bond Lengths in Organic Com- pound s. Cornell University Press. L. Orbital Interactions in Chemistry. Schaefer Ill (ed . New York. and / or physi- cal ch emistry. and Whangbo. Plenum Press. W. F. Burdett. (1985). Vol. Draw the virtual orbitals tha t were not show n in Figure 1. Unified Valence Bond Theory of Electronic Structure. Chem. 3d ed. NY. New York . 3.-H. Orpen.. (1975). En glewood Cliffs.8 for the planar and p yra midal forms of CH 3 and in Figure 1. John Wiley & Sons. Academic Press. Electronegativity. If there are two types of ligands b ound to a trigo nal bipyram id al metal a tom. N. G. Qualitative Molecular Orbital Theory.. and Goddard..64 CHAPTER 1: I NTRODUCT I ON TO STRU CTURE AND MODELS OF BONDING 32. W.D. III in Modern Theoretical Chemistry. II. K. M. T. G. (A classic book that is unfortunately out of print-ge t o ne if you ca n!) Valence Bond Theory.. H. The Nature of the Che111ical Bond and the Structure of Molecules and Crystals.. D. A. Ithaca.. A. "Tables of Bond Leng ths Determ ined by X-ray and Ne utron Diffraction. T.. Academic P ress. and Resonance Pauling. Resonance Theory in Organic Chemistry. The Organic Chemist's Book of Orbitals. Selected references to more advanced topics are given below. F. Perturbation Theory. Berlin. New York. New York. yo u shou ld consul t you r introdu ctory texts in gene ral che mistry. John Wiley & Sons.1. (1979). NJ .9 for the linea r and bent forms of CH 2 . W. G. J. L. (1995).

Gibbs free energy.e. We also explore the structures of some of the "exotic" molecules that chemists have made in an attempt to push the limits of structure and bonding. such as eclipsed vs. We then show that the energetics of basic organic molecules can be estimated using a surprisingly simple model called thermochemical group increments. While this simple method has limited theoretical justification. 65 . a logical first topic is variou s trends in bond dissociation energies. To start off. This leads to the idea of strain energy.. such as in cyclopropane. a very important concept in physical orgaruc chemistry. enthalpy. Since strain and stability can be tied to bond strengths. Next. the effects on reactivity are left to the chapters in Part II of this book. These in- clude topics such as aromaticity. or propose a potential syn- thesis of a new target. In this discussion. and discuss the fact that all internal motions of molecules are dictated by internal forces that are represented by potential surfaces. but also the weaknesses. and the goal of this section of the chapter is to present the basic tenets and the strengths. we combine our discussion of sta- ble molecules and reactive intermediates in order to emphasize the similari ties and the differences. As in Chapter 1. we link stretching vibrational modes to bond homolysis. of the method. a major goal of the chapter is to explore the energetic consequences of deviations from the standard geometrical param- eters described in Chapter 1. leads to the answer. In ad- dition. The main goal of this chapter is to bring the student "up-to-speed" in this thought process. and the varia- tions that occur with ring puckering.e. Finally. and carbanions. and entropy). staggered ethane. and the quantities used to measure them (i. we extend the discussion of internal motions and consider more subtle variations from standard structural parameters. dramatic deviations from standard bonding parameters. This is a field known as conformational analysis. we will show that the relationship between deviation from standard bonding parameters and energy can be put on a quantitative basis using the molecular mechanics method. s tabil- ity. often associated with increased stability. However.. Almost all practicing organic chemists now avail themselves of this method. the anomeric affect. The focus is on the energetics associated with structure. carbocations. and others that can be rationalized by orbital interactions. a rapid evaluation of strains and stabilizing effects. in part. making it a must topic for organic chemistry in the early 21st century. we review some basic concepts of thermochemistry (i. In that regard. We also consider molecules in which special bonding arrangements lead to novel structural and energetic consequences. We end our discussion of thermochemistry by considering basic reactive intermediates- radicals. it is remarkably use- ful in predicting the structures and energies of a wide range of organic structures. Too often this method is used as a "black box" computational tool. chemists routinely use the heat of formation as a number that can be compared from structure to structure. strain and stabil- ity). and conformational effects all have ramifications on reactivity. are covered. or attempt to predict the lowest energy conformation of a new struc- ture. except for a few Connections highlights that emphasize our current discussions. CHAPTER 2 Strain and Stability Intent and Purpose When organic chemists consider the reactivity of a new molecule. the overall stability of a compound is defined. Next. Strain.

1. however. such as the force created by an explosion of TNT. A stretched spring has more internal energy because it can do work by returning to the reference state. Importantly. or a d ifferent conformation where the strain is relieved. and systematic varia- tions are seen that clearly relate structure to energy. such as a spring. We refer to the diagrams as potential energy diagrams (or surfaces). absorption of a photon of infrared radi- ation can lead to a h igher energy molecule that has an excited vibrational mode (see a brief discu ssion below). Chemi- cals can do work via chemical reactions. along with strain. just as w ith a spring. are relative. In that sense. but are quite telling when compared to the same number for another molecule. or a brick raised above the ground. Furthermore. However. This results in m ore internal energy within the strained molecule relative to the reference. or absorption ofUV light can lead to a higher energy electronic state (see Chapter 16). and we comment upon them in this ch apter where appropriate. as an analogy to internal energy can be quite informative. chemists often write reaction coordinate di- agrams and conform ational analysis plots with the energy axis labeled as PE or just E (Fig- ure 2. there is also thermal . To obtain a strained chemical s tructure or higher quantized energy state. where the spring now has stored potential energy (PE) (increased internal energy. but there are many ties between the two concepts. and we present correlations that show how useful the energies are in comparing similar structures. where that energy is capable of being released. Molecules can also occupy different quantized en ergy states associated with variou s in- ternal vibrational and electronic states.1. strain energy being only one example. there must always be a refer- ence state. All thermod y- namic values. The reference structure may be a completely different chemical compound that lacks the particular strain. These phenomena also add potential energy to the molecules. or can be placed in the molecule during a synthetic procedure.1 Thermochemistry of Stable Molecules Structure and energy are intimately related. Hence. Very often. Fundamentally. It is important to always know w hat the reference is. energy needs to be added. examining a common everyday structure.1 A) that can be released by returning to the relaxed form. we are leading up to a precise definition of a key concept in organic chemistry-strain energy (see Section 2. u sing Gibbs free energy or enthalpy (see below). For example. and these should be consid- ered. If the relaxed state of the spring is the reference (the most sta- ble arrangement). The energy can come from a collision.66 CHAPTER 2: STR A I N A D STAB ILI TY 2. Chemists do not typically associate these higher energy states with s train. see Figure 2. perhaps pulling a block in the process. In this chapter we discuss the types of energies involved. Because of this analogy to potential energy. 2. certain bond- ing arrangements lead to a stabiliza tion of a chemical structure. The introduction of strain into an organic molecule is perfectly analogous to the stretch- ing or compressing of a spring. it is an al- ogous to potential energy within a compressed spring. For organic molecules. In organic chemistry. all energy is related to the ability ofa system to do work. What is internal energy? It is the energy held or stored within a m olecule. a large number of experimental energies of different kinds are known. when analyzing a molecule's structure and reactivity.1 The Concepts of Internal Strain and Relative Stability What do chemists mean by strain? We are referring to a structural stress within a mole- cule that is not present in some reference compound. strain is typically associated with a conformational distor- tion or nonoptimal bonding situation relative to standard organic structures. then stretching or compressing leads to a strained form of the spring.1). in order to fully understand the strain that is being discussed. This concept nicely conveys the notion of higher internal energy for one structure on the diagram relative to another. which implies some function that constrains the different internal motions of the molecule or any possible chemical reactions. we are more explicit in defining the energy. Associated with any structure is a " total en- ergy" or "internal energy"-numbers that are of minimal value in isolation. Part of this en- ergy can be released when given an outlet such as a chemical reaction.7). from a photon.

in the spring upon compression Relaxed Relaxed spring spring Extension and compression axis Strained transition > state Cl iiir::: Q) Potential energy (also called B. .. If the energy re- quired to obtain the higher energy state or strained form of a molecule is lower than or com- parable toRT. let's recall a few of the basics of thermodynamics. except now B is more stable than A.1 THERMOCHEMISTRY OF STABLE MOLECULES 67 Compressed spring Potential energy stored A. B. In summary. less stable. Various quantized energy states. 2.: :: -. Chemists use several terms to describe this situation.--- state -~~* coordinate diagram for A +± B. This will be important when we examine small barriers to confor- mational changes. toRT (where R is the gas constant and Tis the temperature in kelvin). on a per mole basis. C. The Boltzmann distribution (covered in Chapter 7) tells us how to predict this based upon the temperature. which is measured by temperature. such as rotation of the methyl groups in ethane. Compressing a spring leads to increased potential energy (internal energy) in the spring.. The goal here is to gain an understanding of the forms of potential energy that chemists use to discuss chemical stability and strain.----. when we place a molecule higher on a potential energy diagram. then this state will be significantly populated at that temperature without the addition of more energy. and I or strained. can be populated depending upon the temperature. We compare the energy required to get to the higher energy state.0 Structure A \~. A reaction > c. Before looking at some types of strain or stabilizing factors. -~Grxn o a Structure B Reaction or conformational analysis coordinate energy in the solution. stating that these structures are higher in energy. ~ strain energy) required to c achieve a transition between ! A and B 0 a. ~ U) . !:> Q) r::: Q) Q) __rj-. Structure A Structure B Reaction or conformational analysis coordinate Figure 2. or strained chemical structures. we are implying that it has more internal energy than the more stable reference compounds placed lower on that plot. analogous to the picture of a spring.0 .1 A. Analogous diagram for a Strained transition chemical transformation between A and B.

A transform ation that has a ~Go of-1.5. 2. Different terms ("exothermic" and "endothermic".2 designate concentration. the ~Go of a transforma- tion is best thought of as the difference in stability of two different compositions of an ensem- ble of molecules at standard states and at constant pressure. the transformation of A to B is exer- gonic.68 CHAPTER 2: STRAIN AND STAB ILITY 2. This has an important consequence. 2. 0 The free energy has two components. Gibbs Free Energy It is the change in Gibbs free energy between two different chemical states (compounds or conformations) that determines the position of the equilibrium between these states.1). K eq' and the free en- ergy change for any chemical process. the analysis focuses upon a chemical reaction. 2. see Section 3. we always examine the relationship between two or more states.1) [B] (Eq. Typically. 2. Conversely. 2. and entropy in entropy units (eu). Eq. Figure 2.1 C shows a reaction coordi- nate diagram for the interconversion of A and B. In this regard. for en- tropy to be an energy value. (Eq. where B ha s less internal energy. the enthalpy (~H ) and the entropy (~5°). In this chapter we consider only thermodynamics. The unit kJ /mol is also commonly u sed. and the conversion is 1 kcal/mol = 4. a spontaneous change in the composition of the solution will occur to create B. where we show that Gibbs free energy is akin to a driving force (or a potential energy) for a spontaneous change in composition. A useful relationship to remember is that every 1.0 kcal/ mol at 298 Khas a 85:15 ratio of Band A.5 for a discussion of standard states and insight into the origin of Eq. measured in mol /L.2) Keq = [A] Recall that the square brackets in the equilibrium relationship given in Eq. If we start with excess A. It is also important to remember that when we say" a spontaneous change will occur" we are implying nothing about how long it will take for that ch ange to occur. Eq. The same equations are used to describe a full-fledged reac- tion or a simple interconversion of two conformations. We are u sing kcal/ mol as the standard unit of energy in this text. which are equi valent to cal/mol • K.1 presents the relationship between the equilibrium constant.36 kcal/ mol in ~Go is worth a factor of 10 in an equilibrium constant at 298 K (see Table 2. so we are concerned with the relative energies of A and B-the ~G • In Chapter 7 we will discuss the p a thway in terconverting A and Band the 0 meaning of ~G*. when the free energy of B is higher than A.1. ~Go (the o symbolizes we are considering standard states. the reaction is end- ergonic. or a difference in population of quantized energy states. 2. related by the Gibbs-Helmholtz equation (Eq. Changes in temperature affect the free energy between A and B.1. Therefore. We 0 denote the Gibbs free energy change for any process as ~G • Here. enthalpy is measured in kcal / mol. a conforma- tional difference. Our experiments tell us the ex- tent to which one state is preferred over the other. but at -78 oc we will see a 93:7 ratio . Eq.1). Use these terms when examining a Gibbs free energy diagram. we need kinetics (Chapter 7) to pro- vide a time scale for the transformation.184 kJ/mol.1 tells us that temperature effects are important also. see below) are u sed when the energy axis is enthalpy.2 Types of Energy When we compare the internal energy of one molecule to another as a means of un- derstanding stability and strain. it must be multiplied b y the temperature. and therefore . 2.1. Thermodynamics only tells us the direction in which the change will occur.2 gives the ratio of species A and Bat equilibrium for a simple transformation that interconverts A and B (A <=± B).3). When the Gibbs free energy of B is lower than A. 2. A more precise definition is given in Section 3.

. But.3.9 9. In this chapter we will explicitly define several forms of strain: bond angle strain. we say the reaction is endothermic (t:.4. the easiest way to view a change in enthalpy is to associate it with changes in bond strengths. and a few examples of this will be given in Chapter 16. the energy of a system also changes due to variations in temperature.H 0 is positive).).36 10 90.x °. and we will explore bond strengths extensively in this chapter. where now the energy difference between two states has been transferred in the form of heat. Each of these strains results from the weakening of the bonds in the molecule.33 70 30 -1.Ho is negative). hydrogenation.5 2.1. If the conversion requires an uptake of energy.0 13. 2. 2. Chemical reactions often perform work. showing how an equilibrium constant is influenced by temperature.1 and Eq.2 93 7 the equilibrium constant. For exo- thermic and endothermic reactions we place the energy of Blower or higher than the energy of A. we obtain Eq. In relatively rare but interesting cases. in the laboratory. The energy be- tween a starting and final state changes as work is done. we say the reaction is exothermic (t:.15 K ( -78 °C) -1. Hence. (Eq. such as combustion reactions of gasoline that drive the pistons in car motors. 2.1 -2.H0 ). 2. In fact. When the conversion of A to B releases energy. respectively. etc.. so instead the reactions release or absorb energy by changing the tempera hire of the reaction vessel. but normally just written as t:. We will look at several heats of reaction in this book (such as heats of formation.08 1000 99. This is what is most relevant to chemistry. the energy is released in the form of a photon. Combining Eq. For any reaction or change in conformation we define a heat of reaction (designated t:. 1 T HERMOCHEMISTRY OF S TABL E MOLEC U LES 69 Table 2. That is why this chapter is primarily focused upon enthalpy considerations. most often in the form of heat. 2 . which can cool the solution. on a reaction coordinate diagram (or conformational analysis plot) with enthalpy as the energy axis. combustion. We have already stated that the energy of any system is defined as its capacity to do work. most chemical transformations are not set up to do work.9 0. + eq RT R Enthalpy The change in enthalpy is defined as the change in heat between two different composi- tions of an ensemble of molecules at constant pressure if no work is done. 2.4) InK =.72 100 99 1 -4.15o (Eq.0 5. etc. torsional strain. all kinds of strains and issues of stability can be related to bond strengths. a change in heat is accompanied by a change in bonding between two states (both intramolecular and intermolecular bonding).1 -0.. This value reflects the difference in heat between 11 two compositions of a solution.3) LV{o . In chemistry.H. which makes the molecule less stable.1 Composition of an AlB Mixture as a Function of Gibbs Free Energy Difference and Temperature ~G o K %B %A At298K 0 1 50 50 . However. steric strain.44 85 15 At 195.

or a bond stretch occurs. high frequency vibrations (as found with strong covalent bonds) make no significant contributions to the entropy of the mole- cule. 2. By configu- rations. Lastly. becau se they can display large variations among closely related structures. This means that the loss of translation of a molecule during a reaction will cost around 30 eu. Especially when differences in ~Ho are small. rotational. the greater (and more favorable) the entropy. The relationship is more complex for configurations that are not at the same energy. There is also a statistical viewpoint from which to examine the entropy of a system. the more favorable the entropy. The more freely that a bond rotates. and so we will dis- cuss entropy much more in those chapters. Examples of this are given in the following Going Deeper hig hli ght. convincing arguments about the relative stabilities of two or more structures can be made by considering ~Ho values alone. Here we refer to every kind of internal motion of the molecule. we mean in part different geometries-for example. Vibrational en- tropy is much more complex. The more configurations that a system can exist in. For conformations of the same energy. This is a gas phase reaction. Howeve1~ we must always remember that it is ~Go that sets the equilibrium constan t. The easiest way to consider entropy in chemis try is to associate it with all the kinds of molecular and atomic movements. In general. a bond angle bends. entropy effects are often very important in molecular recognition and solvation phenomena.3 Bond Dissociation Energies We have already stated that enthalpy is best considered in chemistry as being manifest in bond strengths. and vibrational. 2. In addition. we may have to consider entropy. ~5°. Recall from introductory organic chemistry that BDE is defined as ~Ho for the process shown in Eq. referred to as degrees of freedom . the more favorable the entropy. There are three different kinds of degrees of freedom: transla- tional. respectively. the difference in entropies. all species are in their standard states. For this reason.2 on ring vibrational modes). The rotational entropy is also high. and a wide variety of techniques has been developed to determine BDEs directly or indirectly. the lower the entropy because the movement of the molecule is more restricted (see Section 3. The more concentrated the solution. the more favor- able the entropy. Internal bond rotations contribute on average around 3 to 5 eu . which gives the entropy differences between linear and cyclic alkanes. It is independent of concentration. These are convenient numbers to remember. The more conformations a molecule can have. and various forms of bond angle vibrations. There are approximate values th at we can give to the entropies of certain degrees of free- dom. the majority of the discussions of thermochemistry 0 focu s on ~H • Often. The translational entropy of a small mole- cule at 1M concentration is around 30 eu.1.3. the entropy increases as Rln(n) where n is the number of identical conformations. as we will see in Chapters 3 and 4. Increased heat al- ways leads to more molecular disorder.1. such as bond stretches. will be fairly small.5 for fur- ther discussion). because they can give quick insight into entropy changes that occur between two states. around 30 eu for a common organic molecule. but low frequency vibrations can contribute a few entropy units (see Section 2. bond rotations. the more kinds of motions and the more unconstrained those motions are. A valuable thermodynamic quantity that ha s been extensively studied is the bond dissociation energy (BDE). and the more " loose" these are. often when comparing two similar structures.5) . conforma tions (see the discus- sion of conformational analysis given below).5. This means that the restriction of a freely rotating bond during a chemical transformation will cost around 3 to 5 eu. 2. While entropy is certainly important for significant changes in chemical structure (such as a cyclization). R-R-R"+R" Ml 0 = BDE (Eq. and the bond cleavage is always homolytic (producing two radi- cals).7Q CHAPTER 2: STRAIN AND STAB ILIT Y Entropy The entropy of a system is a measure of the disorder of that system. The BDE provides a rigorous definition of a bond strength. Translational and rotational refer to the translation of the molecule throughout space and the tumbling of the molecule. The more degrees of freedom. the more disordered it is.

Sci. "Entropic Contributions to Rate Accelera- tions in Enzymic and Intramolecular Reactions and the Chelate Effect. If yo u want to just remember one number when considering how much entropy is lost when a bond rotation is frozen. and this weakens the bond by decreasing orbital overlap. compar- ing the BDEs of various bonds allows us to consider which are the strong and weak bonds in a molecule. the valence orbitals of X get progressively larger. The CH3-X BDE trend sup- ports this.0 eu change for the freezing of a bond rotation.8 Page. the trend is F > Cl > Br > I. the values are temperature dependent.1683 (1971 ). Nat/. 1678.. and depend upon whether they are measured in the gas phase or solution (Table 2. most chemists use 4. and C. w hich is stronger than the analogous C(sp3 )-H bond. Data su ch as these lead to the general view that shorter bonds are stronger bonds. We will also examine the relative enthalpies of conformations /'-. as we move down the periodic table.halogen bonds.0 -21. Note that the entropy difference becomes less favo rable as the ring gets bigger.D of cyclic systems.. . Hence. Similar trends are seen in H-X bond strengths: 0-H > N-H > C-H. Since BDE is a meaningful representation of the strength of a particular bond. ~ -. 0 . We no ted in Chapter 1 that orbital mixing decreases with smaller overlap and larger energy gaps.. BDEs are reflective of the nature of the chemical bond..7 -. Electronegativity is the major influence operating here. linear and cyclic. -. 2.2 gives gas phase values for room temperature). and Jencks. For CHrX bonding. At right are shown a series of differences in entropies between linear -1 3.9 ~ that leads to the conclusion that there is on average a 3 to 5 -. M. in that the larger the electronegativity difference.L -7.3 /'. W. As with all thermochemical data. there are clear trends. P. A C(sp)-H bond is stronger than a C(sp2 )-H bond. N. In addition.. even though C and I have comparable electronegativities. In considering the methyl. 1. cyclohexane is an exception to the gen- eral trends in cycloalkanes. This is especially true of thermal and photochemical reactions." Proc.. Table 2. and several trends are evident. Acad. Hybridization and resonance also contribute significantly to bond strengths. so our models of chemical bonding should be able to rationalize variations in bond dissociation energies. with the exception of cyclohexane. the stronger the bond . We noted in Chapter 1 that introduc- ing some polarity into a covalen t bond can strengthen the bond. the first being F > OH > NH2 . which is the least favorable rela- ti ve to the open chain form. In addition. Recall also from Table 1.. CH3X. it is the comparison of the entropies of these two kinds of systems. The larger orbital size leads to a size mismatch with the carbon valence orbitals.2 we'll show later.19. and the pref- erences for particular conformations of sing le bonds. and cyclic alkanes. USA.4 that bond lengths follow the inverse trend : C-F < C-0 < C-N < C-C./'-. assum- ing there are no large entropy changes. 68... more bond rotations must be frozen to form the ring.5 eu. Although we must always keep in mind the distinc- tion between thermodynamic and kinetic stability.. The same electronegativity effect discussed above contributes to this trend. As the rin gs become bigger. This is the main effect in the first-row series involving F. in which bond homolysis is often a key initial step. 1 T HERMOCHEMISTRY OF STABLE MO LECULES 71 Going Deeper Entropy Changes During Cyclization Reactions In this chapter we will examine the enthalpy changes for !lS 0 (eu) bond rotations in linear chemical structures.2lists a number of prototypical BDEs. The sum of all the BDEs in one molecule relative to the sum in another molecule with the same number and types of bonds will tell us which molecule is more stable. However. As ~ -. -10. in many instances the thermodynamic BDE can be u sed to predict the reactivity of a molecule. the C-I bond is much weaker than the C-C bond.o Entropies of cyclization .

H 98 .H 90. so we only b riefly review it here with one exam p le.6) CH3. An interes ting contrast is see n with th e peroxides of Table 2. As described in an upcom ing Going Deeper highli ght.2.. G.6) C6 H5 CH 2-H 88 (89. c(CH 2)4 .4) CH3. 6) CH3. Som e interesting BDEs are associa ted with bonds to oxygen.SiMe3 89. We w ill see thi s effect in practice when we study radi cal reaction s in Chap ters 10 an d 11. 36. reconci l- in g va riations am ong resu lts obtained fro m di fferent methods. In contrast. let's con sider BDEs in an absolu te ra ther th an a re lati ve sense. M. if it translates to a comparable d ifference in !Gneti c s tability.5 ) CH3.2) c(CH 2)5.NH2 84..1) CH3._~ o -H HO-QH 51 CH3. The roughl y 50 kca l / mol difference in ther- m od yn am ic stability.H 94 (96. The reactants possess th e stronger bonds. leading to this trend in BDE.5) I CsHs-GsHs (118) CH3.H 86 (88.7) CH3.. w e ca n use bond strength d iffer- en ces between reactants and produ cts to predict the exothermicity or endothermicity of a reaction .C2H5 1 (89.-----------.H 11 0.H 132 (131.9) CH3.------. and Ell ison. 493 (1982). Before leaving this topic.H 94.8) CH3...C(CH3lJ !(87..I 57. F. Th e 0-0 bond s of p eroxides in general are quite weak.r--------J c(CH 2)s-H 95 .6) CH3. 2. 33. We only need to look at the d ifferences in the BDEs for those bonds th at change between reactants and products.SH 74 .9) CH 3. --------·.. B.3 (88.8) CH3..3 CH3C(O).1) CH3. The impor tan t point is that we ca n . thi s has sub- stanti al consequences in biology.5 CH30 .H 1044 (104.PbMe 3 57 I ( )H 73 H2C= C H 2 (174.2) CH2= CH.H 1074 (107..---··---··---··-·..·-·----···- (C H3)2CH.5 NH 2..H 11 0 {11 0. 1) .. S.1 (98.0CH 3 I (83.2 {101 .C 6 H 5 (103.H 96 . This is an exercise routinely done in introdu ctory organi c chemis try classes.3 CH 3S. Blanksby.H 11 9 (11 8. Clte111.7 (87.7) CH3.j. Consider the reaction given in Eq.2 (96. di- me thy l p eroxide has a BDE of only 38 kcal / mol.9 (72.2 Some Specific Bond Dissociation Energies (in k caUmol) * Bond BDE Bond BDE Bond BDE~ H.CH(CH 3h 1 (88.1) CH3-QH 92 .6.6 H2C= 0 (178. whjch means they are the more sta- ble s tructures. -·---·---·----] (CH3lJC.l C H3." A1111.H 104. 2 ( 57 ~ c(CH 2lJ. making this typ e of bond very su sceptible to thermally induced hom olysis. 0) I H J [[>.CI . and hence the reaction is endothermic.72 CHAPTER 2 : STRAI AN D S T A BILI TY Table 2. this makes alkanes in credib ly stable.CH 2C 6 H5 (77. N um be rs not in pa re n theses a re from M cMillen.SnMe3 71.. D.1 (105. would cor- resp ond to a factor of 1037 d ifference in reactivity! Using BDEs to Predict Exothermicity and Endothermicity Since bond strengths are measured in enthalpy units.C and C-H bonds generally have BDEs well in excess of 85 kcal / mol.4 I . An 0-H bond is one of the stron gest covalent bonds.8) CH3.Br 70.H 106.5) CH2= CHCH 2.6 (38) CH3.5) I __j *The bond of in te rest is shown in color. Res.2 I 82 .255-263 (2003).· .H 90 .0) C6 H5 . RcE Pltys. As shown in the following Connections highlight. 2) I I I h 974 HC = C H (230.5 HO." Ace.H 93. _____. Clte111.H 105.SiH3 88 .GeMe 3 CH 30 -QCH 3 37.F --·-- CH 3CH2.1) .3 (92.H 86. Va lues a re from two sources.4) CH3. C.7) CH3-G H3 904 (90. 6) l CH3-G H=CH 2 (101.~.1) HC = C.· .1) ~=~]. and Gold en.9 (85. Numbe rs in pa rentheses a re from a recent attempt to provide the most current estimates.1 o -H HOCH 2..9 (11 2.. D..H 95 ..CH2CH=CH2 (76.2 (104. " Hydroca rbon Bond Dissociatio n Ene rgies. Rem ember from Ch apter 1 that more s character in a hybrid orbital makes the group more electron ega tive and d ecreases the bond length. "Bond Disso- cia tion Energies of O rga nic Molecules.

H bond abstraction of a H from a C-H bond in a deoxyribose moi- of water. predict whether a reaction is exothermic or endothermic simply by examining which side of the reaction arrow has the stronger bonds. S. F. Normal modes are the vibrational degrees of freedom that a molecule has-namely.!RT . x ra ys. HO• . In Chapter 7 we will introduce the Arrhe. Exposure of a li ving system to high energy rad iation (-y rays. Greenbe rg.2 (Eq. (1995). with a BDE of119 kcal I m ol. and over time. aro und for at m ost only 1010 years. and. The m ost common source of biologi- ca l h ydroxyl radi cals is ioni zing radiation. H ydro xyl radical does indeed react The Hydroxyl Radical in Biology rapidly w ith DNA. Water is ple ntiful in a biologi. different workers have produced detailed analyses that choose the best value for a given BDE.H + H . 2. d1emistry caJl ensue. Often. ety. all beca use of the great strength of an cal system. a major reaction pathway being One of the strongest bonds in ch e mi stry is the 0 . too. This creates a potentially lethal (to the cell) les ion in ics often presages reactivity in radical chemistry. it should the gene tic m aterial. the trends in BDEs are the same for the two. high should be exothermic for almost any organic compound energy electrons) leads to ioniza tion of wate r and ulti- RH. are therm ally very stable! stants from acti vation energies: k= Ae-E.1 + 119 92. Where comparisons are available.1. That is. Chapm a n & Hall.. hydrocarbons nius equa tion. and it is generally very fast. CH 3. cess that can genera te h ydroxy l rad ica ls from water to is extreme ly reacti ve.OH for h omolysis of ethane at 25 oc would be approxi- the activation energy for the the rmally induced homolysis mately 1044 years.lfHO • were to form. bends.4 An Introduction to Potential Functions and Surfaces. The structures of molecules are not Going Deeper The Half-Life for Homolysis of If we assum e an Arrhe niu s pre-exponentia l factor (A) Ethane at Room Temperature of 10 13 (a common va lue for a unim olecular process).(92. perhaps most Vol. S. some quite lethal chemistry are profound .90 kcal I mol C-C BDE of ethane sets a lower limit to half. including HO • . but the generation of HO• is a hi gh energy OHbond. Our uni verse is postulated to have been of the molecule. New Yor k. From that point on.0H + H. it would rapidl y abstrac t Foote. there are small but significant differences. which can be used to calculate ra te con. and indeed this is the case. As s uch. etc. Structure Energetics n11d Reactivity in Chemistry. Since thermodynam. 2.3 + 104. 2. dif- ferent techniques produce slightly different values. the homoly tic cleavage of a bond that defines the BDE is a chemi- cal reaction that can be related to a normal mode within the molecule.2. J. We present there- sults of two such analyses in Table 2. p rocess. Va le ntine.2) = 27.. 2.6) f. proteins. one from 1982 and one from 2003. J. C. though. In general.H 105. w hen comparing two different BDEs it is best to use values from the sa me data set. A. Thus. Certainly.1 + 119) . tor- sions.. we would expect any pro- no t be surpr ising to hear that the hydroxyl radical. Each has its own fundamental frequency. Thi s Active Oxyge11 in Chemistry.1 THERMOC H EM I ST RY OF STABLE MOL E CULES 73 Connections A Consequence of High Bond Strength: significantly. a reaction such as have severe consequences fo r a biological system.3 + 104. a hydrogen fro m almost any organic molecule. the The . mately the production of man y toxic oxygen-based spe- The impli ca tions of this observation for biological cies. stretches. includes carbohydrates.Ji o = (105. and Liebma n. wags.Bond Stretches For a simple molecule.6 kcal / mol A variety of experimental techniques has been employed to m easure BDEs. DNA.

At very sho rt dis- tances between the atoms (small ron this curve).2 shows several such plots. bends. Population of a high er energy vibration leads to w h at is called an excited vibrational state of the molecule. The lowest energy vibration (n = 0) is d efined as the zero point en- ergy (ZPE). Each fundamental frequency is described by a potential function. noting that they arise from internal forces that cause bonds to form (see Sec- tion 14. These functions all derive from the bonding forces that hold the atoms together in the molecule. which are destabilizing. l:<1 0 :t <1 Continuum of ---. The bonds stretch and contract. A Morse potential is a plot of energy (specifically enthalpy) as a function of distance between the atoms. where vis the frequency of the bond vibration (this equation is actually for a harmonic oscillator. and conformational changes.7. Figure 2. The rungs represent the quantized vibrational sta tes. but are instead quantized as given by Eq. the Morse potential.74 CHAPTER 2: STRAIN AND STA BILITY A. a mathematical function that d escribes how the potential energy changes along the coordinate for the internal motion. The shape of this curve is indicative of an anharmonic oscillator. We do not show many of these mathema tical functions. Because the atoms vibrate. The energies of the vibrations are not continuous. but instead has the ZPE. and the atoms are repelled apart. and nuclear repulsions. As the bond becomes weaker. At very long atomic distances (long r) the atoms move freely with respect to each other. and we show this in a Going Deeper highlight on page 76 after having d efined force constants. which describes bond- stre tching vibrations. Chapter 14 discusses the origin of the shape of these curves. such as bond stretches. absorption of a photon of light corresponding to the energy difference between the states can form the basis of a spec- troscopic method (see the discussion of IR spectroscopy given below). The energy minimum represents a balance between the overlap of atomic orbit- als. One result of the vibrational energies being quantized is that the bond never actually has the energy represented by the lowest point on the Morse potenti al surface. c. As w ith any series of quantized energy states. Bonds vibrate in a manner constrained by the Morse potential. the Morse po tential becomes more shallow and the difference between the energies of the vibrational states decreases. For m ost compounds at ambient temperature the vibrational states for bond stretching above the ZPE state are not significantly populated. Let's start with one of the simplest potential functions. called potential surfaces. instead. but are dynamic. Higher energy vibrations are the o ther rungs drawn on the Morse potential. which will be a very important notion that we will return to when isotope effects are discussed in Chapter 8. which is stabilizing. B.'r-t-rT- o -----r--="~ infinitesimally spaced energy states BDE First four stretching vibrational ZPE energies Figure 2. Chemists pictorially tie the strength of a bond to its length using a Morse potential. with correspondingly lower frequency ZPE v ib ra tion .5). but instead we simply draw graphs of the functions.1. 2.2 Morse potentials for bonds of varying strength. the nuclear repulsions dominate. static. The goal is to give a qualitative insigh t into how potential functions d escribe each and every molecular motion. see the next Going Deeper highlight). their . A to B to C represents increasingly weaker bonds. The energies are drawn on the Morse potential as rungs.

The balls are moving most rapidly at vibrations. Every spring nected by a spring. instead of an anharmonic oscillator (Morse poten. For the low energy vibrational states. The energy of spend most of their time at the short distance and long the v ibration changes due to increases in the amplitude distance. ljf/ starts to resemble a continuum on the extreme ends of extension and compression Break in Break in the PE surface the PE surface w Q. the aver- tial). Just as with electron wavefunctions (Section the average distance between them. Let's explore this using the potential probabilities in the second quantized state. The sec- function that describes the electron motion. The second diagram shows the at Particular Separations same picture except with the square of the wavefunctions. On the changing direction at the extreme compression and exten- rungs are drawn the wavefunctions that describe these sion of the spring. because the balls slow down when they are of the wave that describes each vibrational state. The probability same procedure is used for bond vibrations. w hat is The lowest energy vibration has the highest probable the probability of having a particular bond length?" This bond length centered right at the traditional bond length is similar to questions related to the probability of finding (ro)· These are the bond lengths reported in Table 1. Square the wavefunction to get probability of distances 2 The harmonic oscillator. We are using this potential to what is observed in our macroscopic world for springs. the next highest.2). During a vibration. One may ask. However. the two balls possesses its own fundamental frequency. w Q. anharmonic oscillator that describes real bonds. showing ljf and ljl . as energy increases the number of nodes (zeroes in ability of finding them at this distance is low. We squ are is high at short and long bond lengths relative to r 0 . Let's look at the probability of a potential surface that describes a particular motion. and we draw the quantized energy states The square of the wavefunction finds distances between again as rungs. potential. the surface for a harmonic oscillator (such as with a normal average distance between the atoms is still r 0 • With an spring). the Morse potential is our first example thus indicating probabilities. which we will probability function decreases toward the short and long show in Ch apter 14 is related to the square of the wave. the fre. energy level (see Figure 2." At any given moment.1).4. As already stated. 2. This is exactly w hat is found for two balls con- quencies of the vibrations do not change. The exact ond vibrational state is quite interesting. of bond lengths for three different vibrational states. Shown below is a potential surface for the vibrations Lastly. and a very high energy state. The electrons at particular coordinates in space.1 T H ERMOCHEM ISTRY OF STABLE MO L ECULES 75 Going Deeper The Probability of Finding Atoms the function) increases. For both bonds in low energy vibrational the atoms that are most probable at sh ort and long bond states and for springs in the macroscopic world . lengths. model bonds. bond lengths confined by the potential surface. the high energy state shown starts to approach of a harmonic oscillator. the The bond vibrates within the constrain ts imposed by this ZPE. the harmonic age bond lengths actually increase with each increasin g oscil lator nicely mimics the anharmonic oscillator. This the wavefunction that describes the wave-l ike nature of mean s that there are actually two bond lengths with high the bond vibration. and hence the prob- 1.

than n. (Eq. and u sing Eq. As a bond gets weaker. 2. The value of v from this equation is u sed in Eq. Imagine the extreme of a bond stretch-if it is large enough the bond just brea ks. The shape of the Morse potential is indicative of the strength of the bond. the Morse po- tential is more shallow and the width becomes greater. Bonds beh ave in much the sa me way for the lower energy vibrations. That ga p gives a preference for . and the vibration is considered to be a loose vibration (Figures 2. Now the frequency of vibration is lower.8 gives the vibrational frequency of a spring with two masses (m 1 and m 2 ) at the ends. and the loose- ness of the associated vibration (hence. the concept of d escribing a vibration as stiff or loose has consequences for the en- tropy of the compound. 2.8 to calculate the C=C in an alkene 11 fundamental frequency. and the entropy of reaction. This equation gives the frequ ency.7 bond stre tches. For otherwise similar bonds. 2.2 Band C). = (m 1m 2 ) I (m 1 + m2)]. meaning th at onl y for Bond Stretches at T = 298 K? a sm a ll fraction of the molecules ex ist in the 11 = 1 energy Previous ly we commente d that vibrational states higher leve l. Using this force constant in Eq. and is di scussed qualitatively in another Going Deeper high- light on pa ge 75. the energy that need s to be put into the bond to fully break it (the BDE) is shown as an arrow that s tarts at the bottom rung (ZPE) of the curve and ends at the reference state with the atoms at a n infinite distance. Note that a larger k gives a larger frequency to the vibration.5 bond. As described in the above Going Deeper highlight. = 0 are not important for exa mining b o nd dissocia- Force Constants for tion e nergies at ambient temperatures. which we will return to in Chapters 7 and 8. 2. where enthalpy is defined as zero. We kn ow thi s by Type of Bond a Stretch (mdyne/A) calculating the energy difference between the sta tes w ith 11 = 0 and n = 1. A deeper potential represents a stronger bond because the BDE is larger. just as implied in Eq.5 for the reaction that defines BDE.22 kcal / mol.7) It m akes good sense that the bond stretching vibration is the internal motion within the molecule that leads to homolysis. 2. wh ereas a loose spring v ibra tes slowly.76 CHAPTER 2: STRAIN AND STABILITY positions move with time relative to each other. The force constant (k) for the spring reflects the strength of the bond. where m.C C-C in an alkane 4. its entropy). v.2 A). A narrow potential surfa ce indicates a stiff vibration.7 to calculate the energy of the s tre tching vibration . Below we lis t a table of force constan ts for C-Hi n a ty pica l me th yl 4. Eq. Hence. Recall from your everyday experience that the fundamental frequency of a stiff spring is large. meaning that the bond length varies with time. where stronger bonds have larger force constants. The manner in which one predicts the probability of finding a particular bond length requires quantum mechanics. m eaning that the potential surfaces are more narrow (Figure 2. As noted earlier. A very similar equation is used to model angle b ending motions. for a harmonic oscilla tor. 2. The energy axis is rela tive to the free dissociated atoms.8) Going Deeper How do We Know That n = 0 is Most Relevant the low energy sta te of 233 to 1 at 298 K. 2. the trends in vibrational states can be easily rati onalized by modeling a bond as a spring that connects the two a toms. we obtain 3.7 to calcul ate the C=C in an alkyne 16 energy difference between the 11 = 0 and 11 = 1 vibrati ons. en == (n + t) hv (Eq. is the reduced massofm 1 and m 2 [m . the stronger bonds have higher vibrational frequencies. Let's focu s o n the force constant fo r a C.

and while it can be produced by ab initio. it is not of much value to practicing chemists. we have the standard heat of formation. it is not known in an absolute sense.1 v= 1249 cm. respectively. the typical frequen cies for C-C. the force con- stants must increase in this series (as shown by the table in the previous Going Deeper highli ght). / H A~ ~ +'-. molec- ular wags. th e total energy is the energy required to convert the molecule into isolated atom s (breaking all bonds) and then remove all electrons from those atoms.1. Recall our discu ssion of enthalpy.9-the fo rmation of the molecule from its constituent elements. 2. the heat of formation. where energy was related to the ability of a system to perform work or release or absorb heat.1 Figure 2. we therefore need a reaction th at can release or absorb heat.5 Heats of Formation and Combustion For any m olecul e. Furtherm ore.H0 for the process of Eq. Figure 2. For the purposes of this textbook. In theory. A more useful energetic quantity associated with a particular molecule is its heat of for- mati on. If the elements are in their standard states (graphite for C. and scissoring motions are also quantized. + 0 0 . it is simply important to recognize that all vibrational degrees of freedom are quantized. For example.1. H 2 at 1 atm for H). there are only certain energies that these other vibrations can have. Bond angle bends. alon g w ith the IR absorption values. total energy is relative to some standard state taken for the isolated atoms. and 2260-2100 cm-1. The heat of formation of a hydrocarbon is defined as t. 2. Bond s tretch es are not the only vibrations that are quantized. giving insight into th e relati ve entrop y of one vibration to another. symbolized as t. and hence excita- ti on fro m one energy level to another also gives absorption bands in theIR spectrum. 2. 1617-1640 cm. and a reference state to which we can relate all molecules. This is typically a very large number (tens of thousands of kcal / mol). quantum mechanical calculations. Because the reduced masses are all the same.3 IR frequencies for the deg rees o f freed om of form aldehyde. For example. To define the energy of a molecule.1 v= 1167 cm.1 v= 2843 cm. t 0 <~'H-.1 v= 1746cm. C = C. all m olecules are referenced to the elements. However. and are found at smaller cm-1 in IR spectra. and that each moti on is defin ed by a potential surface that describes it as stiff or loose.Ht All enthalpy values are typically expressed on a per mole basis.1 0 Ot • 0 0 HX H ' H~ H A 0 H G HGJ " CH scissor 2 "' CH 2 "' rock CH 2 wag v= 1500 cm. any reaction that all organic molecules can undergo could be used. There are separate potential functions and surfaces that describe each of these vibrational modes.H + H_) CH2 symmetric stretch CH 2 asymmetric stretch CO stretch v= 2783 cm. If other elem ents are involved in the molecule. chemis ts have settled upon one particular reaction. Furthermore. H ence. which are defined to have a heat of formation of zero. then we include the elements in their . Just as with stretching vibra- tions.1 TH E RM O CH E MISTR Y OF S TABLE MO L EC ULES 77 Infrared Spectroscopy The trends discussed above can all be observed in infrared (IR) spectra. These vibrations ty pi cally are at lower frequency. This comparison shows that the stronger bond has the higher frequ ency vibration. as with all energies. and C= C bond stretches are 450-500 cm-1.3 shows all the normal modes (vibrational degrees of freedom) of formal- dehyde.

a larger negative heat of formation means that the compound is more stable. 2. all f. the ener- bending motions of a CH 2 group in ethylene and methane gies of the out-of-plane bending motions are quantized. The reaction that creates the more stable organic molecule from the elements releases more energy.78 CHAPTER 2: STRA IN AND STABILITY Going Deeper Potential Surfaces for Bond Bending Motions toward other C-H bonds. 2. let's compare the out-of-plane face as shown). and hence this bend tial surfaces. One mus t also designate a temperature.Hr values will refer to 298 K. (Eq. thanks to the petroleum industry-the heat of combustion has been determined accurately.Ht for C02 and H 2 0 are known (-94. It is less sterically demanding to All internal degrees of freedom can be described by poten.10-the "burning" of the molecule. 2.05 and -57. As with all degrees of freedom. stretching vibrations placed on a Morse potential. When compar- ing two or more structures that are isomers. The out-of-plane bend rungs in a manner analogous to the energies of the for ethylene sends the hydrogens toward the 1r bond.9) Heats of formation are typically not easy to mea sure directly. it is easy to convert a heat of combustim~ to a heat of for- mation.)H 2 (1atm) (Eq. 2.9. The correct designation is then t. and 0 unless explicitly noted otherwise. This is defined as f. which are considered to be loose or stiff. and so there must be less internal energy within that .10) Students often confuse the trends in heats of formation and combustion. but typically we will assume 298 K and not write it. while the same bend in methane sends the hydrogens Out-of-plane wag Out-of-plane wag PE PE \ Hydrogen angle deviation Hydrogen angle deviation from the xy plane from the xy plane Out-of-plane bending motions standard states on the left-hand side of Eq.Ht 298 . and this is how most experimental t:. As in ethylene is easier (described by a looser potential sur- just one more example. 2. Since t.Hr0 values are actually obtained. nC (graphite) + (-'q.9 repre- sen ts a reaction that is difficult to perform and isolate a single product! However. (we do this in preparation for the discussion of secondary which could be written on these potential surfaces as isotope effects given in Chapter 8).H 0 for the reaction of Eq. respectively).80 kcal/ mol. for many organic molecules-especially for hydrocarbons. Certainly Eq. push the hydrogens toward a 'Ti bond.

Thermodynamic stability involves the position of a par- ticular equilibrium. Table 2.CH2-CH2. There- fore.81 4. That is. knowing the heat of formation of a molecule tells us the position of the equilibrium of Eq. Clearly. relative to the elements. kinetic stabili ty.82 kcal l mol. that is. It is best to see how this happens by just looking at the patterns.G* (see Figure 2.93 CH3-CH2.92 .31 CH3-CHz-CH2-CH2-CH3 -35.24 CH3-CH2-CH3 -24.89 4.96 4.3 Values for the Heats of Formation of Simple Linear Alkanes (in kcal/mol) Structure CH3-CH3 -20. We know from experience that there must be a very large kinetic barrier to the conversion of benzene to its elements.3 shows experimental t. The t. 2. a mole of cyclohexane is more enthalpically stable than the equivalent amount of graphite plus H 2 • A further insight arises when we compare cyclohexane's heat of formation to another molecule with the same molecular for- 0 mu Ia.CH2-CH2-CH2-CH2-CH3 -44.G* must be very large (going from right to left) for the "reaction" of Eq.5 kcal l mol). and reactivity. 2.G 0 . the compound is releasing its internal energy in the form of heat. and so it is set by t. definite patterns arise. but rather some standard organic molecule that we can compare other molecules to. our reference state in effect becomes not the elements C and H. Ki- netic stability involves the rates of reactions. Assuming for the moment that entropy effects are small.25. A molecule is kinetically stable if there are substantial activa- tion barriers to all available reaction paths. however.9 lies well to the right for cyclohexane and well to the left for benzene. Keep in mind that this comparison must always be tied back to the ele- ments. From these patterns we can d e- velop several notions that are very valuable in considering the thermodynamic stabilities of organic molecules. Let's consider the heat of formation of benzene plus three H 2 (t. we would conclude that the equilibrium in Eg. more strained.6 The Group Increment Method As already stated.1 C and Chapter 7). 2. as emphasized earlier in thi s chapter. Heats of formation become useful when we compare values for similar molecules.G 0 . benzene is thermodynamically unstable relative to its constituent elements (graphite plus H 2 ) .Hr0 values for a series of n-alkanes.85 CH3-CHz-CH2-CH2-CH2-CH3 . Many heats of formation values are available. so it may surprise some that t. and compare it to cyclohexane. In contrast. but by t.CH 2-CHrCH 2-CH3 -49.1 THERMOCH E M ISTRY OF STABLE MOLECULES 79 organic molecule (the molecule is less strained). For example.60 CHrCH2-CHrCH3 -30. m ethylcyclopentane is Jess stable than cyclohexane. a large negative heat of combus- tion means the compound is more unstable. What is the meaning of this number? The fact that it is negative means that Eq. not something that is of great value to most organic chemis ts.9.Ht values become even more powerful when we compare molecules of different molecular formulas . t. such as methylcyclopentane (t. because it is kinetically stable. the heat of formation of cy- clohexane is -29.00 4.9 is exothermic.9 kcal I mol.1. 2.84 4. Then. the more heat is released when it burns. and it is worthwhile to spend a few mo- ments to consider some differences among them. Benzene does not spontaneously convert to graphite and H 2 .96 CH3-CH2. and so is controlled not by t. 2. The more unstable the structure.9 when benzene is the or- ganic molecule. After some initial Table2.39. In other words.Hr = . When we start to compare heats of forma- tion for a number of hydrocarbons.H t for H 2 = 0 kcal I mol by definition). Benzene is an especially "stable" mole- cule.Ht for benzene is +19. We must always keep clear in our minds the difference between thermodynamic stabil- ity. In combustion.15 5. 2. cyclohexane is actually the more stable structure.

93 kcal l mol to 6-Ht This pattern continues.Hr0 values for a wide variety of organic molecul es with a con- siderable diversity of functionality have been determined. Note the mann er in which a group increm ent is desig- nated.86) = . and these are given in Table 2. not Figure2. References to more complete collections are given at the end of the chapter. 6.(H) = 5 (3. and hence.93 kcal I mol. There are examples where the Benson group values fail. and it gives surprisingly good results. When this is done and averaged over a number of alkanes. there is a clear p attern.20) = -20.10. These group in crem ents are sometimes referred to as Benson increments. some bonding parb1ers are implicit. Much less data are ava ilable for such com- pounds. This allows group increment val- ues to be established for many types of groups. we state a few general observations.20 kcal l mol) . We will use the group increment method sporadically in later chapters. The symbolism for the CH 2 group we have been discu ssing is C-(Hh(Ch-a carbon attached to two h ydrogens and two alkyl carbons. shown for isobutyl benzene. The process is straightforward.51) = 5. in which a group such as CH 2 also contributes a transferable set of orbitals.8Q CHAPTER 2: STRAI N AND STAB ILI TY variation. and suggests the no- tion of a group increment-a consistent contribution to the 6-Ht of a molecule by a particu- lar grouping. and a few cavea ts.H t usi ng gro up 1 C 8 -(C)= 1 (5. 1 C -(C 8 )(C)(Hh = 1 (. Only this kind of a CH 2 has a group increment of -4. however.90) = -1 .4lists a small subset of all available group increment val- ues. Table 2. _ 2CH 3. because there are two CH3's.Hr is predicted to be exactly (n. the quality of a group increment is only as good as the quality of the experimental thermodynamic data th at were used to obtain it.93 kcal I mol is small. There are enou gh experimental determinations of 6-Ht for alkanes that group incre- ments for an alkane CHand an alkane quaternary carbon (C) can be derived. design ated as C-(Hh(C).4 shows h ow to calculate 6.Hr becau se it is attached to one alkyl carbon and one carbonyl carbon (although the devia- tion from -4.4. As an exa mple.93 kcal l mol) for the four CH 2's.4. and so the group increments are quite reli- able. 6-Ht values for a wide range of molecules to be estimated.51 increments.86 1 C-(H)(C) 3 = 1 (. However. which sometimes contribute to hea ts of form ation. First. a va lue of -10. such as diazenes and oximes. who was th e prime developer of the concept. Averaging over a large dataset reveals that every addi- tional CH 2 group adds -4. Fur- thermore. for example. The CH 2 of 2-bu tanone makes a different contribution to 0 6.93 kcal l mol) + 2(-10. We can now calculate a priori 6-Hro for 0 any linear alkane. For the basic hydrocarbon groups. it is simple to determine the group increment for a CH3 group in an alkane. 15 kcal/mol Experimental : -5.50 Sample calcu latio n of t:. and so considerable caution is in order w hen u sing these group increments. see below) . for example.4. vast quantities of thermodynamic data are available.4 5 C 8 .Hr0 for isobutylbenzene. and divide the remainder by 2.1. and soon we will introduce som e additional fea- tures of the method. so that is n ot designated. a Cd (an olefinic C) mu st have a cd parmer. Figure 2. the Benson group method ign ores interactions between groups. At thi s point. after Sidney Benson.90 2 C-(HlJ(C) = 2 (.20 kcal l mol is obtained for the CH3 group increment. We take 6-Ht for n-hexane. Similarly. For CH 3 (CH 2) . 6.enough attached atoms must be given to unambiguously id entify the type of group. subtract 4(-4. Once we have a CH 2 group increment.2)(-4. On the other hand.30) = 16.40 -5.34 kcal/mol . This notion of a transferable energy increment associated with a chemical group is nicely consistent with the bonding model discussed in Chapter 1. there are published group increment valu es for some fairly obscure functional groups.15 ± 0.

000-gallon scale! In more typi cal ph ysica l organj c cherrustry. in part because the entrop y differences between m olecules are often sm all.5 C-(O)(Cu)(H)z -8.08 Cd(CO)(C) 7. we noted that bond rotations and other low energy molec- ular motions contribute to the en tropy of a m olecule. however.6 C-(Cs)(C)(H)z -4. in chemical engineering. 0 Along with group increments for 11Hr there are analogou s group increments for !1 St In .1.2 C-(O)z(Hh -16.(CO)(C)z C-(Cu)(Ch 2.6 N-(CO)(C 6 )(H) 0.6 CO-(Cd)(H) -29.8 C. .78 C-(CO)(C)3 1. C11 = benzene carbon.5 C-(O)(C)z(H) -7.48 CO-(O)(C) . Both the !1St and hea t capacity group in- crements are applied in the same manner as the 11Ht group in crements.1 CO-(N)(C) -32.9 C-(O)(C)(H)z -8.59 0-(Cd)(C) -30. but because we d o not have enough d ata regarding interactions between d ifferent substituents.35.58 N-(C)z(H) 15.9 N-(C)(H)z 4.(COMCs) -0.1 N.(CO)z(C) -5.9 C-(O)z(C)(H) -16. Since !1G0 determ ines the equi- libri urn constant.08 N. (1976).08 Cd-(H)z 6.4 Cc~-(Cu)( ) 8. S.3 C-(Cd)(C)(H)z -4.10.8 Cd-(Cs)(H) 6.0 NI-(C) 21.8 Cu-(H) 3.30 C-(CO)(Hh -10.9 C8-(C) 5. Thi s is a very important issue. for example.7 C-(Cd)(Cs)(H)z -4..93 Ce-(0) -0.86 CO-(C 8 )(H) -29.3 Cs-(CB) 4.78 O-(C 8 )(C) -23. 1 Cd-(CO)(H) 5.0 N.9 C-(C 8 )(C)z(H) -0.1) .29 CO-(Cs)z -25.6 C-(H)z(C)z -4.20 C-(O)(Cd)(Hh -6.8 Nt-(Cs) 16.3 C-(H)(C)3 -1.c(C)z 10.88 0-(Cs)(H) -37.5 C-(O)(Hh .1 C-(N)(Ch -3. Thermochemica l Kin etics: Methods for the Estimation ofThermochemical Data and Rate Parnm· eters.5 C-(O)z(Ch -18. *Da ta are from Benson. and special tex ts on the topic are available.1 N.34 0-(Cslz -21.3 C-(Cct)z(H)z -4.(CO)(C)(H) -4.1 N.0 C-(N)(C)(Hh -6. they are much less frequently used.4 Group Increments (in kcal/mol) for Fundamental Groupings* Group flHt Grou p flHt Group flHt C-(Hh (C) -10.50 0-(C)(H) -37.64 C-(CO)(C)z(H) -1. h owever.76 CO-(H)z -26.68 CO-(C) 2 -31. and in part because fewer !1St group in crements are available.1 N-(CO)z(H) -18.5 C-(N)(C)z(H) -5.4 N-(C 6 )z(H) 16.9 CO-(N)(H) -29.(CO)(Hh -14.2 C s-(Cc~) 5. Recall that heat capacity reflects the variation of /1H0 w ith tempera ture (see Section 4.(C 6 )(C)(H) 14.6 Cd-(H)(C) 8. In p ractice. our discussion of entropy above. 2d ed.4 C-(Cd)(Ch 1.98 CO-(O)(H) -32.4 C-(O)(Ch -6.26 0-(Cd)z -33. W.0 N-(Cs)(H)z 4.1 0-(CO)(H) -58. and hence it m ay be expected th at alkyl and functional groups would contribute specific entropies.6 C-(CO)(C)(H)z -5. heat capacity group increments are not commonly u sed . 2.81 CO-(O)(Cs) -36. N 1 = imine nitrogen.83 N-(Ch 24. H eat capacity group in- crements are essential when con sidering tempera tures that are considerably different from 298 K.9 C-(N)(Hh -10.51 C 8-(CO) 9. w here it is often crucial to know how much hea t will be libera ted by a particular process-esp ecially if it is being run on a 10.0 Cs-(N) -0.2 0-(CO)(C) -43.4 Cd-(Cth 4.96 CO-(C)(H) -29.7 N-(C 6 )(C)z 26. john Wiley & Sons. New York.1 Nt-(H) 16.5 Cd-(Ct)(C) 8.2 CJ-(Cd)(H) 6.90 O-(C)2 -23. Group incremen ts for heat capacity (11C/ ) are also ava ilable. beca use there is insufficient thermochemical da ta.1 C-(C)4 0.(Cd)(C)z(H) -1.68 CO-(O)(Cd) -32.29 CO-(Cs)(C) -30. !1St grou p increments are potentially useful.2 C.1 THERMO C H EM I STRY OF STABLE MOLECULES 81 Table 2.0 CJ = doubl e bond.1 N.

Ht for cyclo[n] is just n(-4. so it fails for these molecules.7 The problem here is that the smaller rings are strain ed-cyc!opentane to some extent.3 Cyclobutane . acyclic al- kanes. It is tempting to dissect the total strain energy into components. 2. In practi ce.5 shows calcu- lated and experimental t:.2 Thermochemistry of Reactive Intermediates As we did with electronic structure in Chapter 1. Table2. but also some unique features for reactive intermediates. Th ere are many parallels and interconnections. Before looking at thermochemis try. where no bonding distortions occur. We introduced strain at the very outset of the chapter. however. The strain arises. It was parameterized on simple. at least in part.93 kcal / mol).2. in this chapter on molecular structure we will consider reactive intermediates alongside stable molecules. is the notion of stability. indeed.Ht values for simple cycloalkanes..19. and we discu ss all these in this chapter. while cyclobutane and cyclopropane are completely off. the values in Table 2.1 Stability vs. The group increments method suggests that we can predict t:. w e can give a very precise definition of strain energy.1.5 Calculated vs. but one that should be used cautiously.Ht for a molecule and the value of t:. We consider a rea ctive intermed iate to be "unstable". Persistence A recurring theme in physical organic chemistry.Ht calcu- lated using strain-free group in crements. The error for cyclopentane is considerable.Ht Values (in kcal/mol) for Selected Molecules t:.24. and in the next section we will seek to clarify their meaning.7 Strain Energy Now that we have established a definition of the stability of a compound-its heat of formation-we can return to the notion of s train. Now that we have a way to predict the stability of a mole- cule without any bonding distortions. etc. su ch as angle strain. but then things start to go downhill. We describe one structure as stable.7 . because structural issues are fundam entally the same regardless of whether an atom has an octet or not.C angles from the standard value. but another as re- . we mu st consider a general issue for reac- tive intermediates. In fact.18. and cyclobutane and cyclopropane to a much grea ter extent. so often fairly arbitra ry distinctions mu st be made. Observed t:. For cyclohexane the result is quite good. The sam e notions of strain and stability can be applied to molecules even if they are reactive.Hfo calculated by group increments t:. this is a u seful strategy. The group increment method does not account for s train. although for all types of reactive interm ediates there are sp ecial cases that are considered " stable". Strain energy is the difference between the experimental t:.6 -29.7 + 6. this is a dangerous practice.Ht experimental Cyclohexane -29 . These are imprecise terms. One significant result of the group increments approach is that we can now unambiguously define strain energy.9 Cyclopentane .Ht for any organic mole- cule-a powerful tool. In principle. Table 2. and especially in the study of reactive intermediates. from the significant distortion of C-C.7 Cyclopropane -14. The calculated t:. tor- sional strain.8 + 12. 2. however. Let's consider some more examples. ring strain. defining it as an internal stress within a molecule due to some bonding or struc- tural / conformational distortion. The quantum mechanics that defines bonding does not naturally lead to such a p artitioning.4 are often referred to as strain-free group increments.82 CHAPTER 2: ST RAI N AND STABILITY 2.

it is difficult to exactly define what it means to be unstable (reactive). the refer- ence system must be studied under the same experimental conditions as the compound whose stability is being questioned. the contribution to flHt .2. However. however. Chemists often describe a reactive intermediate as "stabilized". Fundamentally. by definition. 2. but for all reactive intermediates the life- time will be extremely dependent on the conditions of observation.11 for a variety of organic molecules. We define a struc- ture as stable or stabilized if it is thermodynamically (Gibbs free energy) more stable than some reference structure.1. stability is an intrinsic property of a reactive intermediate. We will do our best to keep these distinctions dear in the fol- lowing sections.11. there is a vast collection of relevant thermodynamic data for radicals-the collection of BDEs dis- cussed in Section 2. The term is used by chemists in any way that is appropriate to the system under analysis. For example. but only one of which is stabilized with regard to the other. where the compound under analysis is unstable relative to some reference system. we can describe a reactive interme- diate as persistent only with regard to a well-defined set of conditions.2 Radicals More so than any other area of reactive intermediate chemistry. BDEs as a Measure of Stability Consider Eq. a simple variant of Eq. this is an appropriate description of the typical usage of the term "stable" in physical organic chemistry. we must agree upon the definitionof"long" . Yet. What do we mean by this? What does" stable" imply? To be stable. Persistent simply means long lived. the experimental condi- tions affect the reactivity of compounds. under what conditions are we considering stability? Are there any other species present that could react with our" stable" species? An especially lucid discus- sion of this point was provided by K. because as we now describe. we follow the suggestion of Ingold and distin- guish between stability and persistence. our focus is upon electronic stability rather than sterics. Stability is intrinsic to a structure. stability is a thermodynamic no- tion. does a species have to live for minutes.3. thermodynamics is a valuable predictor of radical reactivity. while persistence is very much context sensitive. When comparing Eq. we will in most instances turn to the thermochemis- try of a given situation to predict or rationalize radical reactivity patterns. as discussed above. or days? Furthermore. Similarly. Also. and that will depend on the circumstances. such precision in terminology is notal- ways maintained. and is a term associated with kinetics. both of which have transient existences. In contrast. 2. 2. But in a more practical discussion. because for most reactive intermediates their electronics dominates their reactivity. A persistent structure has relatively high activation barriers for any reaction. 2. A mole- cule that spontaneously reacts all by itself at 0 degrees Kelvin in interstellar space is certainly unstable. while in others we will need a lifetime of days or more before we call the structure persistent. To clarify this situation. There will always be exceptions. Ingold. but it is always used to denote lower internal en- ergy relative to a reference system. Simple variables such as temperature and concentration will also affect the persistence of a reactive intermediate. while persistence is a kinetic one. 2. under typical conditions the ben- zyl cation is not expected to be persistent. Stability implies some sort of minimal lifetime. In some instances a lifetime of seconds will qualify as persistence. and we can never totally ignore kinetics. In addition. Persis- tence is a notion that is very context dependent. U. In the chemical literature. So. leading him to invoke a quote from Humpty Dumpty in Alice in Wonderland: "When I use a word it means just what I choose it to mean- neither more nor less". instability is again always a relative term. we will often encounter two related reactive inter- mediates. Actually.2 THERMO C HEMISTRY OF REA C TI V E I TERMEDIATES 83 active. hours. the benzyl cation is stabilized. For example. a free radi- cal that might be indefinitely long lived under an inert atmosphere could have a very short lifetime in the presence of oxygen. Of course. because it is thermodynamically more stable than its reference. but as a first step. Here.5. This defines the BDE for a generic R-H bond . the methyl cation. Whether a reactive intermediate is long lived or not will depend crucially on whether there is anything around for the reactive intermedi- ate to react with.

2 and focusing on C-H BDEs. with over 10 kcallmol in stabilization fort-butyl radical relative to m ethyl radical. Based upon BDEs. We have to assume that other factors. nu- clear magnetic resonance (NMR) and ultraviolet (UV) studies in 1968 revealed that the ac- tual dimer was the unsymmetrical structure shown in Eg. and it spurred volumes of work. 85. Comparing the BDEs of propene and toluene to an appropriate reference such as ethane suggests resonance stabilization energies of 12. which as with any other process. The clear trend in BDEs of methane > ethane > propane > isobu- tane leads directly to the series of radical stabilities: 3° > 2° > 1o > methyl. There is no trend here. However. This initially controversial result was arguably the birth of reactive intermediate chemistry. Clearly other factors are involved. and t-butyl chloride are 84. BDEs are used to deter- mine the stability of radicals. here. The trityl radical is in equi- librium with a dimer that. 2.1 kcal I mol.0 kcal I mol.4 and 14. which we will explore in depth when discussing carbocations (see below) . Under appropriate conditions. R"+H" t. as we are discussing the relative thermodynamics of a series of reac- tive intermediates. and the stability of a radical is used to rationalize trends in BDEs. Rotating a terminal CH 2 90° out-of-plane completely destroys allyl reso- nance. a good deal of information on radical stabilities can be obtained. (Eq. 2.13. as anticipated from the resonance structures (see Section 1. and so the transition state for rotation is a good model for an allylic structure lacking resonance. for decades.2. BDE va lues are therefore a measure of the relative stability of the radical products compared to the organic reactants.0.1. This assumption actually holds true reasonably well for C-H and C-C bonds. ethyl chloride. such as the stability of R-H (or R-X). However. For allyl radical the rotation barrier has been determined to be 15. R. The reason for this trend is hyperconjugation. the BDE values for the C-Cl bonds in methyl chloride. the relative BDEs for a series of R-H bonds provide an excel- lent collection of relative stabilities of the R • radicals. was assumed to be hexaphenylethane. 2. 2. in acceptable agreement with the direct thermochemical number. 2. respectively.12).7 kcal I mol. BDE is really only the energy it takes to break a bond. We are indeed considering stability. In 1900 Gomberg prepared the triphenylmethyl or trityl radical ac- cording to Eg. Allyl and benzyl radical are substantially stabilized.Ji 0 = BDE (Eq.84 CHAPTER 2: STRAIN AND STABILITY from H• cancels out. vinyl and phenyl radicals are sub- stantially less stable than alkyl radicals. persistence of free radicals is an important one that dates back to the birth of the field.6).11) Returning to Table 2.13. 2. and these BDE values are poor measures of radical stability. and 83. respectively. The overall effect is substantial. Radical Persistence The issue of stability vs. not persistence. 84.3. An alternative way to estimate allyl stabilization is to consider allyl rotation barriers (Eq. not just the stability of the radicals.12) The reasoning behind the above analysis is somewhat circular. are similar in the series for the BDE to reflect solely the radical stability.H. depends solely on the properties of the initial and final states of the system .13) . in which one trityl center added to the para position of a ring of another radical. isopropyl chloride. the free radical p ersists in solution in- definitely at room temperature. As such. (Eq.

Res. The lesson from free radicals that steric protection to rea ctivity is much more important to persistence than elec- tronic stabilization carries over to a wide range of reactive structures. "Free Rad ica l Clocks. sa m- ples of methyl radical have been prepared in which the radical is embedded in glass. Th ese Table 2.14) Figure2.6 refer to fluid solution. radical coupling occurs in an unu sual way that destroys the aromaticity of one ring. \ . For example. and this is why trityl is persis tent. different lifetimes would be observed. D. dim eri zation is completely suppressed (Eq. The extent to which a species is p ersis tent very much depends on the environment. Why.6 The Persistence of Various Radicals* R• t112 . The twist- ing of the rings seriously occludes the radical center. Note that the last two entries in Table 2.. extensive study showed th at the s tabiliza tion provided by the phenyl rings ce nte r is show n in color. the simple C-C bond formation we would expect for other radicals does not happen with trityl. it is not possible for all three aromatic rings of trityl to be in perfect conjugation with the radical center. Once it was appreciated that steric bulk was the key to producing persistent radicals.5 The trityl radical. The radi ca l Ultimately. However. 2.5 shows that the system is quite crowded . 25 oc. \ CF 3 -· --- TMS = trimethylsilyl. U.2 min 8. Che111.3 days . K.5. and three phenyls should provide more stabiliza tion than two. Under other conditions." Ace.. Shown are two views of a calculated stru ctu re for th e radical. with radical concentrations of 10-s M. contributes little to the persistence of trityl. for example). Another clue comes from the ob- servation that placing t-butyl groups in the para positions of all the rings greatly enhances the persistence of the radical. Very long li ved radicals can be prepared without any signifi- cant stabilization. then. a wide range of long lived structures could be prepared. A clue comes from the dimerization chemistry. Trityl would h ave a half-life of ~ 5 min- utes under these conditions. Figure 2. 2. Simple model building will convince you that a fully planar structure would experience severe s teric clashes. at room temperature.4 min ~ I SiMe 3 6 ms Me3 Si-C=C 1 min (Me 3 SilJC· 2. The data in Table 2. but the overall stabiliza tion is not three times the benzyl stabilization. so why does it make the structure more persistent? (+-a 3 C· -~No dimers (Eq . This is indeed th e case. Table 2.. Re- agents cannot easily reach the radical center.6 sh ows a collection of radi- cals and their half-lives under equival ent conditions. 25 oc. I I ~· C-CH > 110 days I \ 1 min Me 3 Si SiMe 3 -q (1-BulJC· 4. In gold. in that both have sp 2 h ybridized radical centers. 317 (1980). 2. Instead . The major factor influencing the persistence of tri- tyl and other radicals is sterics. 1o-5 M -~-· 3 CH • 20 )lS Me 3 Si SiMe 3 -----. Apparently. A para t-butyl group wouldn' t significantl y stabilize th e trityl radical. •cri ll er. 13. In thi s structure. 1 o-5 M R• t112 . . trityl radical is certainly stabilized. The trityl radical adopts the non planar. each ring stabilizes the radical considerably.2 THERMO C H EM ISTRY OF REACT IVE INTERMED IATES 85 We know that a phenyl group stabilizes an adjacent radical due to resonance (benzyl.6 could be viewed as persistent but destabilized radicals. is trityl so much more persistent than a typical radical? Methyl and most other radicals generated under similar conditions would have a fleeting existence. propeller-like structure shown in Figure 2.14).

then the biradi ca l route can be rejected.0 [Cs.2) [C-(C")(Ch ] 1.5 [C-( • CO)(Ch(H)] 2. W. we will ty pically combine in cre- m ents from Tables 2. The m ajor value of free radical group increments is in the prediction of stabi lities of pro- p osed radical and biradical intermediates in va rious thermal and photochemical reactions.8) [Cd-(C • )(H)] 8. H Group Increments for Radicals If we know tlHr0 for R-H.34 [ • N-(H)(Cs)J 38. This is because diffusion is impossi- ble.4. Table2.39. Because we have a large collection of tlHf0 valu es for radicals.. New Yo rk. If it is too high Some stable radicals for the biradical to lie on the reaction path. For example.08 [C. as appropriate. pericyclic process. S. 2d ed.7 [ • N-(Ch] (58. and the methyl radical has no available photochemica l reaction s.51 [ • N-(C)(C 8 )] 42.8 [• C.6 [ • C-(H2)(Cd)] 23. we can derive group increment values for organic free radicals.4) [ • C -( Cs)( C)2] 25. Table 2.( • )(Ch(H) ] -5. For example. and all the caveats concerning group Galvinoxyl increments for neutrals apply even more so for radicals. When steric effects are combined with stabilizing effects such as delocalization. . The symbolism is as before. In addition. N 1 = imine nitrogen.59 [• C-(CMCN)] (56. radica ls that are very" stable" can be prepared .= d ou ble bond.9 [• C-(Ch(Cd)] 24.1 [•C-(Ch(H)] 37.6 [C-(C • )(Hh ] -10. C 0 = benzene carbon.7 Group Increment Values for Free Rad icals (kcal/mol) * Radical tlHt Radical tlH t [•C-(C)(H)z] 35.0 [• C-(Cs)(Hh] 23.5 [C-(•N)(C)(Hh] -6. *Data a re fro m Benson.4 [C 6-N • ] -0. Again.(O • )(C)(H)2] 6. a fairly broad range of group in crements for radicals is available. To calculate the hea t of formation of an organic radical. One valuable piece of information is tlHr0 of the proposed biradical intermediate. a number of spectroscopic techniques h ave directly produced values of tlHr0 for a wide variety of free radicals. Oxygen and any other reactive species cannot com e into contact w ith the radical. {1976}.90 [ • C-(HMCN)] (58.3) [ • C-(C 6 )(C)(H)] 24. or by a concerted.45 [C-(O • )(Ch(H)] 7.1) [Cd-(C • )(C)] 10.8 [C-(C02• )(H)(Ch] . negating all possible bimolecular paths.2 [C-(C02• )(Hh] -47. john Wiley & Sons.95 [C-( • )(Ch ] (-3.82 [C. T!termochemical Kin etics: Methods for t!te Estimation of T!ten noc!tcmica/ Data and Rate Parameters.00 [C-(O • )(Ch] 8. both galvinoxyl and diphenylpicrylhy- .6 [C-( • CO)(C)(Hh ] -0.5 [C-(C02• )(HMC)] -41. drazyl (DPPH) are commercially available free radicals that can be h andled with no special precautions (see margin) .2 [C-(C • )(C)(H)z] -4. then we can derive tlHr0 for R • since tlHr0 for H • is known.C •] 5.(Ch] 38.2) [C-(C")(CMH)] -1. via biradical intermediates. Nevertheless. 7 1ists se- lected results of this effort.86 CHAPTER 2: STRAI N A D STAB I LITY samples are indefinitely stable at room temperature.7 [C-( • CO)(Hh ] -5.7 and 2. Values in pa rentheses are hi g hl y approximate. comparable values for tlSt are also avail- able.5 [ • C-(H)(C)(Cd)] 25..50 [ • C-(H)(C)(CN)] (56. we might want to determine whether an observed thermal rearrangement oc- Diphe nylpicrylhydra zyl (DPPH ) curs homolytically.3 C. and the BDE.0 [ • N-(H)(C)] (55. We will see exa mples of this type of analysis in Chapter 15.

as the latter h as one more carbon. gas phase thermodynamics is an excellent starting point. so anything is an improvement. to compare a 2° ion to a 3° ion in the gas phase. an additional effect mus t be considered. adding "spectator " carbons to a free radical d oes not alter its stability significantly. factors other than cation stability are not as influential as with radicals and carbanion s (see below). and 1-butyl. and it is pretty mu ch saturated by the time we reach a seven-carbon system. Hydride Ion Affinities as a Measure of Stability A common and very valuable m easure of gas phase carboca tion stability is the hydride ion affinity (HIA). Note also that not all the va lues in Table 2. This effect is observed because a naked ion in the gas phase is despera te for solvation of any kind. As such. Before we discuss carboca- tion. but any comparison between ions should take this effect into account. and hydrogen bonding ability of the solvent are important influences. as are the nature of the counterion. adding carbons to a small ion slightly ameliorates this si tuation. Consider the series of 1o cations: ethyl. Adding extra car- bons increases stability at first. in the case of HIAs. we should not compare isopropyl tot-butyl. All introductory courses describe a carbocation stability sequence based on reactivity pat- terns that is 3° > 2° > 1o > methyl. This is less true of other m easures of cation stability.8 were obtained using the same experimental is not desperate for solvation like an ion is.15) HIA values reflect differences in the energies of the initial and final systems under analysis.2 kcal / mol to all system s. Thus. defining the intrinsic stabilities of ion s. and then the effect levels off. It is not a large effect. 2 .8 present many familiar.8 support that finding. And. Just as a larger BDE implies a less stable radical. because H. Before quantifyin g the trend.contributes a con- 0 stant ~Hf value of 34. We did not have to worry grea tly about this issue in our di scussion of free radical s tabili- ties in the gas phase. 1-propyl. stabilities. and the data of Table 2. a larger HIA implies a less stable carbocation. 2. and as such th ere is some uncer- tainty in the numbers. we can obtain ~Ht for the cation R+. and therefore the HIA valu es track very nicely carbocation stability. a theme we will visit many times throughout this book. the trends in the data are quite reliable. however. Any deviation in trends between gas phase and solution studies is likely a consequence of solvati on effects. In effect. A vacuum has the lowest possible dielectric constan t.15. This is simply the heteroly tic analogue of the homolytic cleavage associated with the bond dissociation energy (BDE) just discussed . Ho wever. and subsequently carbanion. It is as if a small amount of solva tion is being provided b y the neighboring carbons. Table 2.8 lists HIA values for representative carbocations as determined in the gas phase. not just carbocation stabilities (see Section 2.2 for sim ilar reasoning related to rad- icals). just as wi th the BDE.2. the usefulness of th e HIA is that it provides a number that can be compared directly for cations of dissimilar structure. Nevertheless. and thus the HIA trends are less sensitive to other factors. If we know ~Ht for RH. . keep in mind that ioni c structures are mu ch more sensitive to environmental influences than radicals. and p erhaps some n ot so familial~ trends.3 Carbocations There is a wealth of information on gas phase ion thermodynamics because of the p ower of mass spectrometry and ion cyclotron resonance techniques. defined as ~Ho for the reaction in Eg. A neutral free radical in the gas pha.2. thermodyna mic information is a less reliable predictor of reactivity for carbocations and carbanions th an it is for radicals. The larger cations are more stable. The data of Table 2. As su ch. 2. (Eq. The polarity. nucleophilicity. The effect is more pronounced in the series of 3° cations derived from t-butyl cation. This pattern is seen in other gas phase ion data. However.2 T HERMOCH E M ISTRY OF RE ACT IVE I N TERMED IATES 87 2. This is in part because the differences between the HIA values for vari- ous cations are much larger than the differences between BDEs for various bonds.

.88 CHAPTER 2 : S TR A I N A N D S TABIL ITY Table2..8 Gas Phase Hydride Ion Affinities (HIA. a nd subsequent years. 246 (!) 225 e /"-.. Academic Press. New York. M./ 247 G 212 )e 231 HC : C-C H~ 270 e ~ 229 H2C=C H 287 e v 228 H2C=C-CH3 258 )e--1 227 Q.-. Gas Phase /on Chemistry.e CH2 234 )e--< 226 Q-\e 220 )ef 224 Q e 287 ~ 231 ~ 249 ~ 225 ~ 230 N H 2 C H~ 218 ~ 218 e HOCH ~ 243 0e 201 FC H~ 290 0 258 N CC H~ 318 *A n excell ent source of gas phase ion da ta is the fo l- lowing series of vol umes: Bowers. (ed. in kcal/rnol) for Selected Carbocations (in kcal/rnol) * CH:}V 312 ~ 256 C2H ~ 273 ~ 236 CH 3 CH 2 CH ~ 266 ~ 225 CH 3 CH 2 CH 2 CH~ 265 ~ 225 re 265 ~ 248 /"-.) (1979).. T.

Aromaticity effects are clear. This is due to the collection of d ata in stable ion media for carbocations developed so brilliantly by Olah and co-workers. though.18). These are very large numbers-significantly larger than analogous comparisons for radicals. the mixin g is a filled orbital with a singly occupied orbital. 3. 2. Substituting just one end of the allyl with a methyl is enou gh to make the delocalized system m ore s table than the localized 2° systems.2 T HERMOCHEMISTRY O F REA C T I V E I NTER MED I ATES 89 Taking the carbon number into effect. The origin of the stabiliza tion due to alkyl substitution was discussed in Section 1. as would other. It should not be surprising that the highly electron withdrawing cyano group greatly destabilizes a cation. and it serves as a model for the intermediate in elec- trophilic aromatic substitution (see Section 10. and 3° ions are more stable than2° ions by about 17 kcal / mol. the four '1T electron cyclopentadienyl ion (see Section 2. and this should destabilize propargyl vs. For reasons discussed in Chapter 3. or equivalently as a hyperconjuga tive interaction. From HIA data. we find that 2° ions are indeed more stable than 1o ions by 18-20 kcal / mol. Consistent with this. while F is destabilizing. In the case ofF. Allyl cation is substantially m ore stable than 1-propyl. similar groups. OH less so. we see that NH 2 is very strongly stabilizing. In the gas phase.7-trimethyl-1-adamantyl cation has been determined. allyl. Many other observations based on Table 2. a substantial amount for an sp2 center. Cations derived from sp 2 carbons are much less stable than those fromsp 3 carbons (based on the instabilities of vinyl and phenyl ca tions). This tracks with the expected '1T donating ability of the atoms.5. Re- call from Chapter 1 that hybridization affects carbon electronegativity in the following or- der: sp > sp 2 > sp 3 . if we remove potential reaction partners we wi ll generally not see dimerization or other reactions in which two carbocations react with each other. Perhaps surprisingly. but is strongly u withdrawing. the 2°I 1o distinction is more im- portant than allylic resonance.l. In a stable ion medium the ions have no possibilities for reactions that quench the ionic charge. Carbons wi th sp h y- bridization are expected to be indu ctively electron w ithdrawing more so than sp2. solvation always attenuates ionic effects. While the gas phase data just described are very informative. it is only weakly '1T donating. we also h ave a significant amount of thermodynamic data on carbocations in solu tion. If we compare the X-CH 2 + ions. As antici. the cationic center lies out of the plane formed by its three neighboring CH2 carbons by a full 0. using ethyl as a ref- erence.5. With a free radical.1 for a discus- sion of aromaticity). The most common reaction . The former model nicely rationalizes why the comparable trend seen in radical chemistry is of smaller magnitude. the substantially diminished stabiliza- tion of the propargyl cation (HC=C -CH 2 +) relative to allyl can be explained as a h ybrid- iza tion effect arising from the electron w ithdrawing alkynyl group. benzyl ion is quite stable for a formally 1o ion. While carbocations are reactive. allyl is less stable than 2-propyl. This relatively sm all gap between a 3° cation and a 2° cation is consistent with the d estabilization due to p yramidaliza tion of the 3° ion.4. as in the much greater stability of the six '1T elec- tron tropylium ion vs. The more electronegative the atom attached to the cationic center. the higher the energy of the cation.212 A. we can estimate the penalty for can be d escribed as a mixing of filled TI(CH3 ) or TI(CH 2) orbitals with the empty p orbital as- sociated with the cation. the x-ray crystal structure of the 3. This structure is simply protonated benzene. Such three-electron interactions are less stabilizing than the two-electron interaction seen with carbocations. H eteroatom effects are interesting.4. so the gas phase data always represent the absolute maximum any ionic effect can show. as in the cyclohexadienyl cation. we find that the difference in HIAs between the 1-adamantyl cation and the isom eric 2-adamantyl ca tion is only 9 kcal / mol. therefore. In support of this.8 are consistent with expectation and experi- ence. 2-Adamantyl pated. As another example. We noted in Chapter 1 that carbocations prefer to be planar at the cationic cen ter. is further stabilizing. We will see in Chapter 11 that experimental differences for reactions in solution that are consid- ered to involve carbocations are nowhere near this large. Similarly. ring 1-Adamantyl constraints prevent the ion from achieving planarity. 7-Trimethyl-1-adamantyl leading to the receipt of the Nobel Prize for Chemistry in 1994. In the 1-adamantyl cation. and the net effect is d estabilizing. Hybridiza tion effects are quite evident. Increasing delocalization.

and others have provided much insight into sub- stitu ent effects on their lifetimes. Richard s. For example.90 C H APTE R 2: ST RA I N AND STABI LI TY paths for a carbocation are capture by a nucleophile and reaction w ith a base to fo rm an ole- fin (see Chapter 11). We can thus envision a stable ion medium . We saw in Section 1. w e are really not making the ion an y more stable. This enabl es spectroscopic studies. Toteva found that the t-butyl carbocation has a li fe time of onl y 10.4. However.counterion.5 kcal l mol in stable ion m edia. consistent w ith the noti on that there is something uniqu e abou t its structure. 2.16. the differe nce in stabili- ti es between 2-butyl and t-buty l cations was determined to be 14.1. carbocations can be long lived. Another ad vantage of these system s is that they are often fluid to very low tem- pera tures. This is a very powerful Lewis acid.H 0 for the reaction in Eq. If we can create an environment that is devoid of nucleophiles and bases. As such. powerful protonic acids such as HF-SbF5 and FS03 H. a typi- ca l va lue for an eight-carbon 3° ca tion. in good agreem ent with the gas phase difference of 16 kcal I mol. but the "stable ion medi a" moniker has stuck. Sb F5 .5 :±: 0. and exp osure of an alkyl halide to it pro- duces a carbocation w ith the Sb 2F10X. Hence. and man y secondary d eriva tives undergo concerted h yd rolysis in w ater that avoid s form ation of the carboca tion reactive interm ediate. We di scussed above tha t a typical2° 13° energy d ifference is 17 kcal I mol.Hr fo r R+ plus Sb 2F 10X.2. A good d eal of evidence in dica tes th at such bridged structures are m ore stable than classica l carboca tions in the gas p hase and in stable ion medi a. Sb 2F 10 X- is a very poor nucleophile and a very weak base. It would perhaps be more precise to refer to " persistent ion media". An interesting analysis ofHIAs can be m ad e for the so-ca lled non-classica l carboca tions. When these were compared w ith comparable d ata from th e gas phase. These data pro- vide compelling support fo r the argument th at w hat is being ob served under stable ion me- di a conditi ons truly is a carboca tion th at is essenti ally devoid of complications fro m so lva- ti on or counterion effects. 2. to be perform ed at low enough temperatures to suppress most carbocation rearrangem ents. The primary 4-methoxyben zyl carboca ti on inter- . Given our discussion in Section 2. Arnett mea- sured both I:!. Ea ch kcal l mol of relative stabilization in the gas phase w as m atched b y a kcal I m ol of relative s tabiliza tion in stable ion medi a. Inert solvents such as S02 or S02ClF can be added as diluents for the SbF5 . the 2-meth yl-2-norbornyl cati on has an H IA of 225 kcal l mol. and it is generally accepted that thi s is a conventi onal ca rbeni um ion.16) In a rem arkable achievem ent. For exa mple. The counterion to R+ is very large. Secondary carbocations are even more reactive toward addition of wa ter. an ex traordinari ly good correlation between the values in the two different en viron- ments was seen. (Eq. In addition. thereb y creating a h ypervalent C. As su ch. As sh own in Table 2.SbF5 allow the generation of carbocations from alcohol and olefin p recursors (see Chapter 5).8. as in the stable ion medium for a series of carbocati ons. allowing th e thermodynamics of ca rboca tion s to be studied. 1 that th ese are compounds with a bridging C-C a or 1r bond to a car- boca ti on center.12 s in wa ter. only 6 kca l I mol hi gher than the analogous 3° system . the lifetimes are extremely short in wa ter. and w e will use it here. we are instead making it persistent. Arnett was able to perform ca lorimetri c measurements in stable ion m edia. Th e substantially sm aller value in this case indica tes a special stabilizati on of th e 2° ion. som e ex tensive studi es by Mayr. direct comparisons of d ata from stable ion media to gas phase d ata (be it experimental or computation al) are valid.16 and th e solvation energy for a series of RX com- 0 pounds. Lifet imes ofCarbocations Th ere are less d a ta related to carboca tion lifetimes as compared to ra di cal lifetimes. the HI A of 2-norbornyl is 231 kcal l mol. although we consider tertiary carbocations stable. For the m ost p art the stable ion media d eveloped by Olah are based on antim on y penta- flu oride. The gas phase HIA d ata support the view th at b ridgi ng impar ts stabi lity. and the negative ch arge is dispersed over a large number of atom s. Yet. In general. 2. they are clearly not persistent in thi s m edium. Thi s allowed a d etermination of I:!. most especially NMR.

5.10 lists anum.0 Ph 3 CH 31. As will be discussed in Chapter approximate. the com. 2.0 acid.17 (in kcal/mol) charge on A. other factors can affect the values.7 Fluoromethane 409 (Eg. Ethylene 44 ene vs. 2.18) PhzCH2 33. 2. Iodomethane 386 mon value reported is the pKa. because s orbitals (unlike p) have finite den- sity at the positively charged nucleus. Selected LlH0 values for the reaction of Eq. as discussed earlier with regard to the use ofBDEs and HIAs to analyze radical *Th e student should reali ze and carbocation stabilities. cyano. hence the 46-unit difference in pKa values for cyclopentadi. because their neutral precursors show a strong preference for concerted hydrolysis that avoids formation of the primary carbocation intermediate.0 pKa = 61 Li (Eq.4 Carbanions By analogy to the previous sections.9. An anion that contains a cyclic array of six electrons is aromatic.0 The analysis of pKa values to make comparisons of carbanion stability is a standard ap. when using pKa values to compare carban- ion stabilities. Cyclohexane 45 (CH3hCH 71 Four electrons is antiaromatic. Under these conditions the carbanion will often be persistent. that the very large values are veat when evaluating pKa values than with BDEs or HIAs.9 PhCH3 41. it is very important to compare the values under as similar a set of conditions as possible.6 mention trends associated with carbon acids with a goal of defining the essential nature of Ethane 420.10 center. with pKa values in the range of 45-70.5 Cyclopentadiene Six rr electrons Cyclopentadiene 16. and sometimes in different solvents some pKa values actually reverse in order. As such. There we will discuss both carbon acids (where the negative for Eq. and each pKaunit represents a factor of 10 in equilibrium acidity.0 Cyclopropene 61 ·· 8 CH3 COCH3 20. and it is often quite straightforward to directly observe carbanions un- der relatively conventional conditions. In such cases. This acid-base chemistry notion is discussed extensively in Chapter 5. Recall (or skip ahead to Chapter 5) that pKa is the negative log Acetonitrile 373. Values of Selected Carbon Acids* stable in a hybrid orbital with mores character. Table2.2 CH3 CN 25. cyclopropene (Eq.6 ber of pKa values for carbon acids.2 THERMOCHEMI S TRY OF REACTIVE INTERMEDIATES 91 mediate of solvolysis of 4-methoxybenzyl chloride exists for several nanoseconds in water. and the effect is even larger with an sp center (pKa 20-25). solvation effects can dramatically change pKa values. Yet. Nitromethane 357. Allyl anion and benzyl anion are greatly stabilized Ethane 50 due to a resonance effect.2. this means that the pKa of the sol- vent must be significantly higher than the pKa of the conjugate acid of the carbanion. Hence. Our starting point for discussion is alkanes.2 CHz = CHCH 3 43 (Eq.84 Cyclopropene Four rr electrons CH3N02 10.19). Chapter 5 is Gas Phase Values of t:.5 Propene 390. Eq. 2. When working in an aprotic solvent. Phenylacetylene 19. a smaller pKa value implies a stronger Acetone 370. a pKa of 50 implies an equilibrium constant of 10-so for the dissociation of the acid! Deprotonation at an sp 2 center (pKa 40-45). Re- member. primarily associated with a carbon) and heteroatom acids.0 Toluene 377. is favored over an sp 3 Table 2. The major requirement for having a carbanion last for a significant time period is for the pH of the medium to be above the pKa of the conjugate acid of the carbanion.19) CH3COCH2COCH3 8. Benzene 400. Electron withdrawing and/ or delocalizing sub. but relatively little is known about the lifetimes of more unstable primary benzyl carbo- cations. Benzene 43 or37 stituents such as carbonyl.1 car ban ions. we should consider Eq.5 proach.5 of the acid dissociation constant in water. .17) Chloromethane 396 Bromomethane 392 A great number of acidities have also been measured in solution.17 to evaluate the stabili.9 ties of carbanions. Table 2. This is a more important ca.17 in the gas phase for carbon acids Ethylene 407. 2.8 are presented in Table 2. A negative charge is more pK. as in ethylene or benzene.Ir entirely devoted to this subject. Carbanion stable ion media exist. CH 3SOCH3 28. Here we briefly Methane 416. 2. Because of the universal importance of acid-base chemistry. 2.18 vs. and nitro produce an expected lowering of the carbon Acetylene 24 acid pKa. Carbanions are generally less reactive than carbo- cations to begin with. 2. 2.

We first explore acyclic systems. For example. these three simple reactions and their associated enthalpies are very good starting points whenever you are considering the relative stabilities of reactive intermediates. wherein molecular mo- tion s directly lead to strained forms. As de- fined in Chapter 1. the paths that interconvert con- formers derive from particular vibrational degrees of freedom. Just as we use heats of formation and combustion to analyze the stabilities of standard organic com- pounds. car- bocations. but the atoms are ar- ranged differently. not on reactive intermediates). the familiar term "sterics" is used. Note that we are not yet considering conventional chemical reactivity. Recall that conformers are stereoisomers that interconvert by rotation around single bonds (see Chapter 6 for definitions of stereochemical concepts). trans. cis. it is time to look at pathways for interconversion between structures. The terminology associated with conformational analysis is in many ways abn to the terminology used to discuss stereochemistry-namely. generally resulting from an accumulation of bulky substitu- ents around a C-C bond. it follows naturally from our discussion concerning strain and stability. Homolysis of C-H bonds is used for radicals and then heterolysis of C-H bonds to create hydride or proton is used to analyze carbocations or carbanions.5 Summary The thermochemistry of the three most common reactive intermediates-radicals. In addition. where the focus is on ro- tations around C-C bonds. strained.2.5). anti. Many of these topics are covered in all introductory organjc chemistry texts. as well as strains introduced upon achieving differing conforma- tions. While this is in some ways an un conventional placement of this topic in a physical organic textbook. We will focus on "sta- ble" structures (i. respectively. So. We will be considering conformers. or conformational isomers. sterics. and so in some places our discussion will be brief. These isomers are not to be confused with constitutional isomers. In Chapter 6 there is a glossary of terms used in conforma tional analysis and stereochemistry. 2. This next discussion will further tie structure and ener- getics together. 2. and vibrational modes that interconvert them. There can be strains inherent to the ring size (small and large). look to the end of Chapter 6 for a definition.1 Acyclic Systems-Torsional Potential Surfaces In considering acyclic systems we distinguish between two types of strain. This leads to bond lengths and angles that deviate substantially from the standard values (we examine some extremes in Section 2. The second type of strain concerns conformers. here we are discussing strained organic molecules in general.3. but the notions we develop here also apply to reactive intermediates. a conformer that is less stable than the ground state is. so if a term is used below that you are unfamiliar with. and carbanions-is conveniently analyzed using heats of very similar reactions.3 Relationships Between Structure and Energetics- Basic Conformational Analysis We will now discuss at some length the many ways in which deviations from standard bonding parameters lead to energetic destabilization of a molecule.92 CHAPTER 2: STRA IN AND STABILITY 2. We begin with the latter type. and their impact on strain and stability. In these discussions. a steric relationship. After considering the acyclic systems. or steric strain. in which the strained form is manifest in the transition state of a rotation or in a less stable conformer.e. In this section we discuss both stable conformers and the transition states that intercon- vert them-a topic generally referred to as conformational analysis. by definition.. syn. The first is distortion of the ground state. are all terms that imply a buttressing of chemical groups that is destabilizing to the molecule. and ties in well with the notion of potential surfaces and molecular motions. we examine cyclic ones. etc. Now that we have covered measurements of stability and strain for both normal and reactive or- ganic compounds. . where the molecular formula is the same.

Ethane Recall the use of Newman projections from introductory organic chemistry. The anti-gauche interconversion crosses a barrier of ca. where the methyls twist relative to each other.4 kcal I mol.9 kcal/mol o~. Thus.6 Left: Torsional itinerary for ethane. Let's 6 X 1010 s-1• This corresponds to a half-life of 10 picosec- calculate h ow fast the bond in ethane is rota ting. We u su ally think of these as the pro- jection angles in a Newman projection.C bond is not" free". 2 . Remember. we use the rapid! Arrhenius equation wi th an A value of 1013 . the eclipsed form is a transition state. rotation about even the simplest C . 3 kcal / mol is quite small. 3.----~----~ oo 60° 120° 180° 240° 300° 360° oo 60° 120° 180° 240° 300° 360° Dihedral angle Dihedral angle H H~H H~H H~H H~H H CH 3 Staggered Gauche Anti Figure 2.----. the staggered conformers interconvert. The x axis of Figure 2. as shown in the next Going Deeper highlight. When the torsional vibration achieves enough energy to traverse the transition state. This is. Right: Torsional itinerary for butane. Ero1--is 3 kcal / mol. the staggered conformation is the only stable form.L----.-----¥~---. 3 R ELAT I ONS H I PS B ETWEE N ST RU C TUR E AND ENE R GE TI CS-BASIC C ONFORMATI ONAL ANALYSIS 93 HH ~ Eclipsed 6 3 Minima H ~HH 3 Max1ma H ~ 0 § "' 3 ~ UJ I 0 . in which one sights directly down a C-C bond and draws the back atom as a large circle and the front atom as a point. As in the onds. Alternatively. it is hindered by a barrier of 3 kcal I mol. Interconversion between one staggered form and another is a torsional mo- tion.C bonds is extremely final Going Deeper highlight of Section 2. showing the three-fold nature of the barrier and the eclipsed and staggered forms. The energy difference between the two-the rotation barrier. The staggered form of ethane is more stable than the eclipsed. h ow com- puter programs evalua te dihedral angles.6 is the dihedral angle between two arbitrarily chosen h ydrogens. However. one on the front and one on the rear carbon . . in fact. the barrier is so small that the rotations are very fast. while the direct gauche-gauche barrier is 5-6 kcal / mol.1. What does it physically mean when w e state that "the staggered form is 3 kcal / mol Going Deeper How Big is 3 kcal/mol? The rate constant at 298 K for a 3 kcal I mol barrier is As an activation barrier. the A-B-C-D dihedral angle is de- fined as the angle between the A-B-C plane and the B-C-D plane. Instead.3. Rotation about simple C.

A contin uum of energy states explicitly discussing quantized energy sta tes for bond exists near and above the top of the barrier. Bond The low e nergy motions appear somewhat harmonic. molecules. M. rotations are just another example of an internal degree of but the torsional motion quickly becomes anharmonic for freedom w ithin molecules. excitations between the energy levels of a tor. in the context of torsions. discussed in Section 2. Phys. why are we not the higher energy states.) and torsional modes implies that a torsion is a vibration. The energy that it takes to undergo the rotation / distortion can come from a collision between ethanes. E. because the other vibrations are quantized. which is 0. Therefore. As the h ydrogens becom e eclipsed. the energy can also arise from a redistribution of the energy between various vibra- tional modes within the ethanes. is less than 3. Here. energy states at or near the continuum. a distortion of the C-Hand C-C bonds occurs (albeit small). Where is this potential energy stored at the transition state? It is placed in the form of a torsional strain. some of the kinetic energy of the ethanes involved in the collision is converted to the potential energy stored in an ethane as it achieves the eclipsed transition state. Clwn. H ansen.. and Dennison.. wags. the rotations appear similar Below we show a schematic representation of this notion to any rotation that occurs in the macroscopic world. states is thermally populated. "The Potential Cons tants of Ethane. D.6 kcal / mol at 298 K. effects on bond rotations are not relevant for conventional As evidence of the fact that torsional motions are organic chemistry at ambient temperatures. Moreover. the energy required for the distortions is released to the ensemble of molecules or into another vibrational mode of the same molecule. Although 0. a large fraction of the lesson s taugh t PE Torsional potential for ethane . G. A signi ficant fraction of the ethane molecules are in ethane are energy states for this torsional motion . Super. but if the tinuum of states for eth ane. the barrier to barrier toRT. and we have already de- fined it as such. etc.1. it makes sense that torsional modes are quantized also. 20. We define a torsional motion as To see if enough energy is present to occupy the con- one methyl group rotating relative to the other. meaning that a collision is not required to achieve the tran- sition state. The Going Deeper Shouldn't Torsional Motions be Quantized? in this chapter about con formational preferences were The bond stretching motions and bond bending motions obtained from the microwave spectra of the compounds. If the continuum of All inte rnal d egrees of freedom are indeed quan tized. it is certainly of th e same order of magni- imposed on a portion of the torsional potential surface for tude.6 the next staggered conformation is not surp assed.326 (1952). This is wh y we tain energies for this single torsion motion are actu ally view bond rotations as nearly " free rotors". As these bonds go back to their normal lengths and angles in the stag- gered form.94 CHAPTER 2: STRA I N AND STAB ILIT Y more stable than the eclipsed form"? As ethane molecules undergo the conversion of stag- gered to eclipsed back to staggered conformations. bends. Only cer. The transfer of energy back and forth between vibrational modes (stretches. How- ever. as we describe in the next Going Deeper highlight. 313.4 only had certain energies. on average 3 kcal per Avogadro's num- ber of molecules is absorbed from the ensemble and then released back to the ensemble. in fact. and quantized allowed . meaning tha t rotations? multiple energy levels are possible. we com pare the 3 kcal / mol energy of this motion is below 3 kcal / mol." sion can be observed in the microwave spectra of organic ]. quantized.

for m ore complex stru ctures the more system atic Klyne-Prelog system is often u sed. and there are two enantiom eric forms on the co m- p lete torsional itinerary.7. and torsion s can exchange energy and be excited to variou s ex tents. 2. now of the methyl groups relative to each other. wags.6 also shows the torsional profi le for rotation about the C2-C3 bond of butane. depending on coll isions and th e tem pera ture. q:.The Gauche Interaction Figure 2.9 kcal / mol. The anti conformer is more stable than the gauche by 0. Thus. bends. The gau che form is chiral.1. as described in the next Goin g Deeper highJi gh t. gauche butane is synclinal (+ or-. both in N ew man projection a nd as the angle between two planes. term ed a vicinal repulsion because it is between groups attached to adj a- cent carbons. th e differing forms of butane p ossess quantized torsional m oti ons. The sta tistica l factor in favor of the gauche shows up as Rln2 in the entrop y (see th e di cu s- sion of entropy in Section 2.gau che barrier. We view the placemen t of the gro up on the fro nt ato m as being in regions of space called anti / syn. In fact. ra ther than crossing the d irect gauche. Left: Definiti on of a dihedral angle. but also to reactive intermediates.7 Dihedra l angles. collisions. for butane w e would calculate the Gibbs free en ergy using Eg.and two types of eclipsed transition states. and d ina!/ peri p la nar rela ti ve to a refe rence group (here a methyl) on the rear atom . The 0. The conven - tion al explanati on of this result is to invoke a steric interaction between the methyl groups in the gau che fo rm. This means th ere is a statistical factor of two favorin g the gau che form w hen considering thermodynamic parameters. 3 R E L ATIO NS HIP S BETWEEN STRUCTURE AND E ERGETICS-BASIC CO NFO RMATIONAL ANALYSIS 95 picture of atomic motion w ithin a molecule that arises is quite complex. 2. (Eq. Th at is. This is a general result of conformational analysis-the easier p ath for interconverting two gauche forms is u su ally through the anti form. 2.4 kcal / m ol above the anti. While gauche and anti are unambiguous terms for butane.20) Just as with eth ane. The transition state for interconverting gau ch e and anti forms is 3. Right: The Klyne-Prelog system fo r describin g conform ations about a sing le bond. depending on which enantiom er). if we want to know the relative proporti ons of ga uche and anti butane at equilibrium. Butane.9 kca l I m ol. anti butane is anti- peri planar. while the direct gauche-to-gauche interconversion must cross a higher bar- rier of 5-6 kcal / mol. or microw ave excitation can lead to torsional energy states possessing the proper energy to overcome the barriers and interconvert the conformers. We could say that gauche butane has a strain energy of 0. The notions of eclipsing and gauche interactions are applicable not only to stable or- ganic structu res.9 kca l/ m ol number is an enth alpy. . c Fig u re 2. and eclipsed butane is synperiplanar. Now there are two typ es of staggered for ms-the gauche and anti conformers. The result is th at butane is roughly 70% anti and 30% gau che at 298 K. Following the guidelines in Figure 2. where stretches.20.2).. the geom etry of radicals is influenced by a conforma tional effect similar to those discussed for stand ard alkanes. Therm al population. we must take into account the fact that there are two gauche forms but only one anti. and of the methyl h yd rogens relative to the propyl group.

Using the values from Table 2. and a few of the most common are given in Table 2. One measurement suggests a barrier of /eri-Bu tyl Radica l. and methyl radical is really the only totally planar radical. there is a conformational effect tha t leads to pyrami- dalization. that's exactly wh at it is. ow consider 3-methylpentane. Draw a Newman projectionsighting down H3c Y cH 3 the C2.Ht values even for Ortho correction 0. the increment should become H considerably larger. the barrier to pyramidal inversio n Padden-Row. However.." f. Clzcm.Ht for 2-methylpentane and 3-methylpentane. so it does not apply to all alkenes. H However. The first such cor- rection is the ga uche correction . Theory substituents. ity at the radical center. K. pyramidaliza- tion occurs in the direction that minimizes ecl ipsing treated as if they were planar.8 = 2-Methylpentane -41. and hence the value given here is a very crude estimate. Soc. the appropri- 3-Methylpentane ate number of gauche corrections should be used.11. Table 2. the best estimate of H tl. the experimental fact is that 2-methylpentane is more stable than 3-methyl- H ~ CH 2C H 3 pentane by 0.11 thoug ht of as a strain energy. H3CH 2c Y cH 3 if one of the olefi n substituents is at-butyl or similar group. a "strain-free" estimate of -42. as developed above. 103. when consid- interactions.36 kca l I mol is obtained for both molecules. So.8 Cis a lkene 1. Once again.ln particular. H the best estimate of tl. M.0 tors have been developed to allow the method to predict accura te tl. and in fac t.C3 bond of 2-methylpentane.0 kcal I mol p enalty is added to 6.76 kcal l mol. There is necessarily at least one gauche interaction. roug hl y 0.36 + 0. H ~ CH 2CH 3 Now the group increment method correctly " predicts" that the 2-methyl isomer is more sta- H3C Y H ble. that the t-butyl is substantially pyrami. under most experimental conditions.66 kcal I mol. a nd Houk. called the ortho correction in Table 2. On considering a large number of structures. Nevertheless./ ~-~-~ . see Exercise 32 at the end of the chapter). there is a subtle ambiguity about how group increments handle Increments (kcal/mol) ga uche interactions.2o H XY L6o H Houk have shown that even the s implest alky l radicals H' are pyram idalized. '\----- 6. Consider calcul ating tl.3°. Similarly. this increment would be group dependent. but by our definition. In each case. Am. Let's see h ow this correction is used. 5046 (1981). the bias is toward dali zed.9 kcal l mol destabilization of gauche butane relative to anti is not typically Table2. So. however. If an alkene is cis. and so is applicable to anti Corrections to Standard Group butane. This increm ent is a fairly crude estimate. a value of 0. In fact. other alkyl radicals. The interpretation of these s tructures invokes simple Pyramidalization of radical centers conformational analysis. but the barrier to pyramidalization is very low. This is a gauche effect. the group incre- ments method is not only used to define strain-free tl. When CH3 using group increm en ts to get the best possible estimate of 6. " O rigin of the Pyramidalization of is quite small.11. H ~ CH 3 mer.Ht for 2-methylpentane using group increments is -42. .12 lists the number of gauche interactions that occur across all bond types.Ht for 3-methylpentane using group increments is -42. ignores such interactions.8 kca l I mol has been settled on as the group increments correction for a gauche interaction.4. A number of correction fac- Gauche alkane 0. with a C-C.96 CHAPTER 2: STRA I N AND S TABILITY Going Deeper The Geometry of Radicals In Chapter 1 we noted that radicals prefer planar struc- tures. N.56 kcal I mol.H r0 for a molecule. Another common group increment correction is for cis double bonds. there really is no strong bias toward planar- predicts. Fort-butyl and the pyramidalization. The group incre- A Few Common ments method.6 strained molecules. There are two gauche interactions-one along the C2-C3 bond and one a long the C3-C4 bond (see margin). not ga uche (in fact. H a 1.C angle of 117.Ht values. . ering the reactio ns of radicals and the potential effects of For ethy l and isopropyl the effect is not large.8) = -40. s imple alkyl radica ls can be The 0.Hr while no such increment is applied to the tran s iso- 0 . At right are sh own three examples.36 + 2(0. So. two substituents in an ortho arrangement on a benzene ring 3-Methylpentane would need a correction. Computational studies by Padden-Row and H''7 H H H ..45 kcal I mol fort-butyl radical.

and the longer the bond. the further the halogen is from the methyl group.1 (1968). In the hal- ogen series. because 3 X 2 = 6. it may at first seem surprising that the larger halogens do not necessarily give larger barri ers.CH2CH3 3. which possess the eclipsing interac- tions.12 Number of Gauche Interactions in Variously Substituted C-C Bonds ForX-Ybond Number of X y gauche R2 CH CH 3 0 CH 2R 1 or2 CHR 2 2or3 R3C CH3 0 CH2R 2 CHRz 4 CR3 6 Barrier Height The barrier to conformational interconversion in butane is larger than ethane due to the larger steric repulsions present in the transition states.4 CH3.7 CHrCH2F 3.2 CH3.CH3 2.0 CH 3 .7 CH3-CH 2l 3. the rotation barriers around CHr Xbonds generally increase as X gets larger.3 CH 3. Phys. as may b e expected. This is because the C-X bond length increases with higher atomic number.BA SI C C O N FOR M AT IO NA L ANA LYS I S 97 Table2. 6. For sim- ple alkyl substitu ents. The C-N bond in nitromethane also has a six-fold barrier. Table 2.NH2 2. Barrier Foldedness The ethane and butane molecules both have a three-fold rotation profile.. J. Org. however.13 shows some other measured rotation barriers around C-C bonds.3 R E L A TIO NS HIP S BETWEE STRUCTU R E AN D ENERG ETI CS." Prog. The barriers to rotation along C-N and C-0 bonds are lower than analogous C-Cbonds. Sketch out a rotational dia- gram if thi s is not clear. Whenever a three-fold rotor (methyl) opposes a two-fold rotor (phe- n yl).1 *Lowe. " Barriers to Inte rnal Rotation Abou t Singie Bonds.0H 1.9 CH 3. on ro- tating 360° about the C-C bond. For example. 2.9 CH3-C(CH3)3 4. the system goes through three maxima and three minima (Figure 2. rotation about the CHr C 6 H 5 bond in toluene produces a six-fold barrier. and the Cp-Co-(COh system (Cp = cyclopentadienyl) can be considered to .CH 2Cl 3. indicating that lone pairs act smaller than C-Hbonds. Not all single bond rotations are three-fold.CH(CH3)z 3. Chem. That is. Table 2.7 CH3-CH2Br 3.6) . w e'll see a six-fold barrier.P.13 A Few CH 3-Y Rotation Barriers for the Bonds Indicated* Barrier height Compound (kcal/mol) CH 3 .

2-Tetrakis(t-butyl)ethane gauche conformer.3-dimethylbutane.H0 = 0! What has gone wrong? H H (H. including a geminal repulsion. Generalizing. One way to rationalize the low barriers is to appreciate that for toluene to have a 3 kcal l mol barrier like e thane but still be six-fold.2-tetrakis(t-butyl)- ethane. the ethane barrier is (3 X 3) I 3 = 3.. and for toluene it is< 0.2.98 CHAPTER 2: STRAIN AND STAB ILI TY have a fifteen-fold barrier (5 X 3). The net effect is that the two forms have essentially the same enthalpies. The anti form has two classical gauche-butane interactions. Erot for nitromethane is 0. 3 interactions repulsion Anti Gauche Figure2.\ Geminal Amplified "gauche" /H.21) A general attribute of n-fold barriers where n is large is that E rot tends to be small. Tetraalkylethanes Consider 2. leading to enhanced gauche butane-type interactions in the anti form only. The consequences of this geminal repulsion are different for gauche and anti-2.006 kcal I mol.3-dimethylbutane exists as a 1:2 mixture anti:gauche. Thus. Thjs geminal repulsion is a general result. The experimen- tal result. As shown in Figure 2..1 kcal l mol.3-dimethylbutane (tetramethylethane). This repulsion counters the fac t that there are more gauche interactions in the ga u che conformer (3) than in the anti (2). however. For ex- ample. an important conclu- H sion of thjs analysis is that all tetraalkylethanes are expected to show a preference for the 1. Thus. the energy of the system would have to change very rapidly in response to only small changes in torsional angle. and these are expected to be destabilizing. geminal repulsion preferentially des tabilizes the anti form.C~CH.8. a steric interac- tion between methyls attached to the same carbon (Figure 2. gem ina l repulsions between methyls ca use expansion of the C-C-C a ng le. is that 2. and a 1:2 mixture (anti:gauche) results from the ~ statistical arguments discussed above. It is true that there are gauche Me • • • Me relationships in 2. the foldedness (F) of a barrier involving an n-fold rotor and an m-fold rotor is given by Eq..2. The net effect is that the two conformers h ave very nearly eq ual enthalpies. For a molecule with a very largeR group. 2. anti and gauche (defined now by the H-C-C-H dihedral angle).8 In tetramethylethan e. further destabilizing that form. where the geminal repulsion does not seriously influence the vicinal repulsion. No such effect exists for the ga uche form. there are two conformers. while the gauche form has three.8). 2.21. Howevet~ there are also other interactions. Thus. in direct analogy to butane.a sta tisti- cal mixture (the gauche form is still crural) implying 6. where q is the number of bonds eclipsed in the transition state. One error in the above analysis is the assumption that the steric relationships present in n-butane will carry over unchanged in 2. The geminal repulsion will occur in addition to the vicinal repuls ion. So.1 .1. such as 1. This rapid fluctuation in energy is an unreasonable situation. the geminal repulsion in th e anti form exacerbates the vicinal repulsion. and the preference for the gauche form of R2CHCHR2 molecules increases as the steric size of R increases.8. .C Q)H repulsions do not lead (vicinal) to amplified interaction \ H3C ~=------CH 3 / \ H3 C CH 3 / gauche H Geminal CH__.3- dime thylbutane. the effect is so large that only the gauche form is viable.3-dimethylbutane. we would expect a 70:30 mixture of anti to gauche. F=(n • m)/q (Eq. As shown in Figure 2.

Bottom: g+g-.structure-forms a nearly cyclic structure (Eq.Pentane Interaction A subtle. However. when reactants are held in a cyclic array.9 Conformations of n-pentane. The g +g. and an adverse H • • • H steric repul- sion develops. In fact. which we consider next. which gives rise to what is commonly called syn-pentane strain. 2.3 RELATIO NS HIP S BETWEEN STRUCTURE AN D ENE RCET I S-BAS IC CONFORMATIONA L ANALYS IS 99 These problems illustrate the danger of quantitatively extrapolating notions of confor- mational analysis based on very simple systems such as butane to more complicated struc- tures.7 is used to differentiate them. A simple but flawed analysis of these two structures would suggest they should be very similar in energy. but we must be wary of their use. Middle: g-g-. A common example is in are- action transition state. This caveat is true both in thinking about structures in general and in applying group increments correction factors. . The g+g- pentane interaction is another important example of the potential vulnerability of simple additivity arguments. The correction factors are the best possible less stable by about 3 kcal / mol.pentane Figure 2. Top: All anti . 2. R~R' =0 R)(R' (Eq. This form brings the two methyl groups very close together. Another important example of this concerns n-pentane and re- lated structures. For n-pentane the all-anti conformer is the most stable.pentane interaction. 2. in some situations gauche interactions will develop.designation of Figure 2. Note the close H • • • H contact in the bottom structure. because both have two ga uche interactions. Recall that there are two possible gauche forms around an internal bond in a linear alkane. but important conformational effect is the g + g.form. As shown in Figure 2. the g + g. Compared to the g-g.9. and the +I.22) g+g. a very significant effect. the ge- ometry is very similar to that of the envelope conformation of cyclopentane (see below). the conformation with consecutive gauche bonds of opposite handedness-the g+g.22) .

The defining feature of cyclopropane is the C-C-C angle of 60°.ln addition. and so an under- standing of their shape preferences is crucial for many studies. Cyclopropane Since any three points define a plane. R~ R' = 0R)(R' (Eg.5 kcal / mol. the C-C-C angles would be 90°. All these observations are consistent with the rehybridization expected in cyclopropane.pentane interaction is allylic strain or A 1•3 strain (also known simply as A strain). but we review and expand upon them here. As shown in Eq. the H-C-H angle is opened up quite a bit to 115° vs. This has an effect on the con- formations of cis alkenes. We just sta ted represents over a 100-fold preference for the lower energy that 3 kca l I mol is a "very signifi cant effect" . H owever. leading to a preference for the more linear structure.pentane interaction). Cyclobutane If cyclobutane were planar. 2.54 A. and the energetic penalty is comparable. but there is also exp ected to be a Cyclopropane significant contribution from the eclipsing C-H interactions across C-C bonds forced by the planar structure.23) allylic strain 2. normal C-C bonds of 1.pentane interaction. The C-C bonds of cyclopropane are short-1.23. The essentials of cycloalkane conformational analysis are covered in introductory organic chemistry. an unfavorable situa- . However. The Au interaction is more commonly invoked-especially in reac- tion transition states-perhaps because only one C-C bond rotation needs to be constrained to achieve the effect (vs. not really amenable to conven- tional bonding models. a value of 106° for the H-C-H angle of the CH 2 group of propane. the only sa tura ted carbocycle for which this is true. Also. in a sense. It is.51 Avs. and an allylic substituent on the other end . the C-H bonds of cyclopropane are more acidic th an normal alkanes. two for the g+g. in Thermodynamics and Kinetics resulting in bond rotations o n ps time scales. To make the C-C-C bond angles smaller. 2. and greater s character lead s to grea ter acidity. conformation. yet a 3 kcal / mol barrier is insignificant we also noted in a previous Going Deeper highlight at ambient temperature. As discussed in some detail in Chapter 14. The s train energy of cyclopropane is 27. the planar con- formation produces perfect C-H eclipsing along the C-C bonds. The maj ority of the strain results from the d eviation of bond angles from their normal values.2 Basic Cyclic Systems The cycloalkanes have long served as important prototypes for conformational analy- sis. This analysis Conformational analysis is a good context in which to highlights a stark difference be tween thermodynamic vs. point out the differences in thermod yn ami c and kinetic kinetic effects. we use more p charac ter. and so here we simply summarize som e important s tructural fea- tures of the molecule. Thi s leaves excess s character in the C-H bonds. A lly lie (A 1 •3) Strain Closely related to the g+g. the bondin g in cyclopropane is unique. a more restricted version of the g+g. A 3 kcal / mol thermodynamic difference phenomena for the same energy value.100 CHAPTER 2 : STRA I N AND STAB I L I TY Going Deeper Differing Magnitudes of Energy Values that a 3 kcal / mol kinetic barrier is extremely sma ll.3. this interaction can develop in a simple ole- fin between a substituent on one end of the olefin. we do not need an experiment to tell us that the carbon framework of cyclopropane is planar. They are also building blocks for many important organic m olecules.

3 RELATIONSHIPS BETWEE N S TRUCTURE AND ENE RGETI CS. Note that the interconversion of envelope and half-chair forms does not require passing through the fully planar form. Inversion of the ring to interconvert the two puckered forms proceeds over a very small barrier (1. g+g. as with cyclobutane. and so the planar form is not a viable conformation. and so in many circumstances one can think of cyclobutane as effectively planar (see Chapter 6 for a more thorough discussion).pentane geometry discussed above. One way is to move just one carbon out of the plane. However. producing the envelope confor- mation. very close to the te trahedral angle. B.10 A).55 A-slightly elongated. These two forms are very nearly equal in energy. An example of where the conformational dy- namics of cyclopentane has important biological ramifications is in the conformations of Cyclopentane half-chair conformation DNA. The rapid interconversion of envelope and half-chair forms makes all hydrogens (and car- bons) equivalent. which has a two-fold rotation axis. the molecule puckers. Note how the substituents (X andY) exchange positions when puckered forms interconvert. or distort this relatively flexible ring to intermediate conformations. most of it due to torsional (eclips- ing) effects. the molecule puckers (Figure 2. This conformer resembles the g+g. Again. A. but again residual torsional strain (the eclipsings are not completely re- lieved by puckering) plays a role. Left: The envelope form of cyclopentane-the conformer conta ins a mirror plane in the plane of the page. producing a time-averaged structure that is equivalent to the planar form. Cyclopentane envelope conformation Alternatively. but now an adverse H• • •H contact is re- placed by a bond (see margin).45 kcal / mol). The strain results primarily from angle bending. The two puckered conformations of cyclobutane. Cyclopentane can distort from the planar form to relieve eclipsings in two different ways. cyclopentane can twist about C-C bonds.2 kcal / mol of strain energy. and they interconvert very rapidly (the barrier is < 2 kcal / mol) by a process termed pseudorotation.10 B). looking down th e two-fold rotation axis. The inversion is sometimes called a butterfly motion. Substituents can cause a cyclopentane to prefer one conformation over the other. planar cyclopentane would suffer from a large num- ber of C-H eclipsing interactions.10 Aspects of cyc!oalkane conformational analysis.10 B). essentially identical to that of cyclopropane. but some due to angle strain. Cyclopentanehas 6. as described in the following Connections highlight. As a result. even though the puckering compresses the C-C-C bond angles (Fig- ure 2.5 kcal / mol. tion . The angle between the two C-C-C planes is 28°. Right: The half-chair form of cyc!opentane. which lies .BAS IC CONFORMATIONAL A A LYSIS lQl y y xV x X A. producing the half-chair form. relieving eclipsing but reducing the C-C-C angles to about 88°. Cyclobutane The C-C bond lengths of cyclobutane are 1.pentane Planar cyclopentane would have a C-C-C angle of 108°. which retains a mirror plane of symmetry (Figure 2. . ~ Envelope -- 'At Half-chair Figure2. Cyclopentane d analogous to _f Cyclopentane vs.5 kcal I mol above these structures. 2. YAYX X ~~- y B. The strain energy is 26.

New York. Cyclohexane chair conformation Like the hydrogen s. A similar interaction is also evident if we sight down the Cl-C6 bond.9 = 1. cyclohexane is generally considered to h ave a strain energy of 0 kcal / mol. Thus. route by which these two forms interconvert are very Cyclohexane Cyclohexane is a benchmark of conformational analysis. The reference book cited b elow has ap proxi- Perhaps the most intensely probed fi ve-membered rings mately 50 pages devoted to just this topic. Cyclohexane is unique in tha t it is predominan tl y found in a single. cyclohexane is much more the exception than the rule for carbocycles. respec- c~N cL~N tively. termed equatorial and axial (labeled He and H .8 kcal / mol.}Q2 CHAPTER 2: STRA I N AND STAB I LITY Connections "Sugar Pucker" in Nucleic Acids complex. Two conformations dominate. Springer. The actual structure is slightly flattened from ideal. Consider a methyl substituent (Figure 2. DNA has 0 0 10 been more extensively studied. or some intermediary form. relatively rigid conformation. the subtle geo- metri ca l issues involving these conformations and the Saenger. we might expect this interaction to destabilize the stru cture by 0.has a profound effect on the global stru cture of the nucleic acid. (1984). In contrast.. Based on the gauche butan e interaction.9 kcal / m ol. If we place a CH3 in the axial p osition at Cl and sight down the Cl-C2 bond. in the picture). In this case. This form is quite compatible with standard bond and torsional angles. From a Newman projection it is easy to see th at each CH 2 contains two differ- ent types of hydrogens. no such gauche interactions develop . the methyl is gauche to one of the carbons of the ring. known as Gr enda (S) the C3 ·-endo(N) and C 2 ·-endo(S).55°. and it is an important building block of many natural and synthe tic products. are the ribose and deoxyribose of RNA and DNA. such that the C-C-C-C torsional angle is . Chair cyclohexane is best appreciated by a Newman projection sighting down two C-C bonds (see below). W. we might expect axial methylcyclohexane to be less stable (or strained) rela- tive to the equatorial conformer by 2 X 0.11 A). The axial and equatorial positions interchange via a ring flipping process (see Fig- ure 2. a substituent can occupy two different positions-axial or equato- rial. the chair form (Figure 2. Ironically. respectively. envelope. The exact conformation of the ring-half-chair. if we place a CH3 in the equatorial position. Principles of Nucleic Acid St ructure. In the chair form.Ve rlag. However. and it can be seen that the 0 differ ing conformations of the deoxyribose ring-termed H the sugar pucker-strongly influence the glob al conforma- tion of the DNA. though. such a simple an alysis .11 A).11 B).

No te that th e chair flip moves axial substitu ents to th e equatorial position and vice versa .B AS I C CONFORMAT I ONAL ANALYS I S 103 A. D . Newm an projecti on s d ow n the Cl-C2 / C5-C4 bonds of methylcyclohexane. In the ax ia l form. B. y 2. Vie ws of cyclohexa ne. Also evident is the s teri c interaction between one methyl hydrogen and the two axia l hydroge ns on C3 and CS. The A value is one m easure of the steric size of a substit- uent. equatorial and axial substituents are in substantially different environments. th ere is a ga uche butane interaction between the meth yl and C3. but an axial substituent puts a decided "kink" in the structure.!lG is called the cyclohexane A value. For most substituents the equatoria l position is preferred . \*yyy y X X X y Gauche No gauche interactions interactions H H B. In th e left s tru ctu re le tters x are ax ial while le tters yare equa torial. A. Interconvertin g ch air forms of cyclohexane. D.14). As shown in Figure 2. producing a very good approximation to the actual situation.3-dimethylcyclohexane in the diaxial conformation. In the equatorial form. A large number of cyclohexane A values ) have been determined (Table 2. and the magnitude of this 0 preference (. the mole- cule retains the overall disk shape of the p arent cyclohexane. C. in that larger groups tend to have larger A values (for another measure of steric size. Note th at chair cyclohexane is a relatively di s k-shaped molecule. with axial and equatorial loca tions labeled. holds up. and axia l methylcyclo- hexa ne.11 C. equatori a l methylcyclohexa ne.11 Aspects of cyclohexane conformational analysis. and an equatorial methyl does little to disrupt this shape.3 RELATIONSH I PS BETW EEN STRUCTURE AN D ENERGET ICS. Cis-1. H~CYR H H0 CH 0 CH 2 3 H v-CH 2~H H v-CH 2~H H H H H Axial Eq uatorial c. Figure 2. In contrast. . an ax ia l substituent puts a "kink" into th e structure. Note the significant steric interaction between the 1.3-diaxial substituents. and in- fluence the overall shape of the mol ecule quite differently.

I will appear to be larger than CH3 .5-2. context is crucial.0.74 CH=CH2 1. the substituents point toward each lar mechanics calculations (see Section 2. New York. Going Deeper Alternative Measurements of Steric Size Taft (see Section 8.68.5 0Si(CH3h 0.42 C2Hs 1.7-4.79 OCOCH3 0. 16.. and they must clash considerably during biphenyl range of structures.75 C6Hn 2.L C H APTER 2: STRAIN AND STA BILITY I 1o4 Table 2. the value of such terms is va r- the rotation barrier in substituted biphenyls. we have the trend CH3 < C2H 5 < CH(CH3 )z < < C(CH3 h However. Nevertheless. when considering the XX apparent steric sizes of simple substituents such as CH3. Ed. the longer bond length of C-I compensates for the larger van der Waals radius of I. see the next Going Deeper highlight).21 Br 0.56-0. Chem.4-2. H. "Steri c Interactions in Organic Chemistry: Spa tial Requirements of ric parameters.14 Selected Cyclohexane A Values (in kcal/mol)* Group A value Group A value D 0. For example. in another context. However.47-0.25-0.50 C02CH3 1.87 Si(CH3)3 2. For an ove rview of efforts to evalua te steri csize.65 CH 2Br 1. given the strong context A common alternative reference system for steric size is effects of steric interactions.3 •Ef iel. Br.60-1. L.5 OH 0. Br. with the ready availability of molecu- to the right. L.006 CH3 1. when clearly there are large differences in their van der Waals radii (Table 1.5).. Wilen. minimizing the adverse steric interac- tion of the gauche form .5-1. appear to be relatively small. Many other attempts have been made to develop ste.2 OC6Hs 0. As such.6) for a wide other. Some substituents with large A values may.2-1. see Forste r.1 COCH3 1. and Mander.48-0.9 I 0. 2.5 N02 1. N. has diminished. able size sequence: Br > CH3 > Cl > N0 2 > COOH > OCH3 > F > H. Cl.61 CF3 2.4 S02CH3 2. and Vag- tie." Angew. F. In addition. Perhaps the best known of these are due to . Stereochemisty of Organic Co mpounds. as shown iable. john Wiley & Sons. universal gauges of the steric demands of Substituents. 429.0-1. substituents.7 CHO 0. (1994).lnt.53-0. we see a perhaps more reason.0 SH 1. and I.8 N(CH 3)2 1.64 CH(CH3)2 2.7 N H2 1..4.55-0. and I are essentially the same.74 F 0. E.1 co2. the usefuln ess of such parameters rotation. This is a good illustration of the fact that with all measurements of steric size. S.79 Cl 0.8 OCH3 0.1).23-1. H.441 (1977). depending on the exact structural arrangement of the sys- tem being used to probe size. However. it seems hard to understand the fact that A values of Cl.04 C6Hs 2. in another structural con- text.21 C02H 1.. In the cyclohexane context. Eng. Here.67 C(CH3h 4. almost any sequence of sizes can be observed. For such a system. some caution is in order when trying to anticipate the "size" of a substituent Q-9 XX Twisted biphenyl in a new context.

For the case of the two methyl groups shown. When two substituents are in a cis-1. This destabilizin g inter- ¢ -~ ¢ ~"~ action is. Soc.3 relationship in a cyclohexane.GHso Hydrogen Bond Strength was determ ined tobe -1. (1994). The energetic consequence of this interaction can be substantial. 2. L..3. A values are frequently additive w hen the axial substituents are trans. When two or m ore groups are attached to a cyclohexane ring. just a cyclic version of the g + g. "Molecul ar Recogn ition of into the A values and the Gibbs free energy of th e hydro. in fa ct.) obtained by combining A values. The definitions can be seen from a view of a flat cyclohexane ring w here the groups are either on the sa me side of the ring (cis) or on opposite sid es (trans). A few examples are given to the right. as d e- scribed in the next Connections highlight. The formation of intram olecu lar hydrogen bonds can H stabilize p articular conform ation s of cyclohexane rings. Cabe ll. let's review some nomencla ture from introductory organic ch emistry that is often confusing.3-diaxial interaction.Diaxial interaction Connections The Use of A Values in a Conformational Analysis The D. C. 2778-2792.BAS I C CONFORMAT IO AL ANALYS IS 105 Before m oving on fu r ther. they can L) ~ ·o . they experience a strong interaction if both are axia l. H H lecular hydrogen bond.3 REL AT I O N SHIP S BETWEE S TR UC T U R E AN D EN ERGE T ICS.14.9kcal/ mol. Form i is destabilized by two axial Evalu ating an intramolecul ar hydrogen bond hyd roxy Is whi le form ii is destab ili zed by an axial isopro- pyl. Using the A va lues of Tab le 2. cis and trans groups can occupy various axial and equ atorial p ositions de- pend ing upon the substitution pattern.A. Chem. Dimethyl cyclohexanes 1 .. H ence. When the A values are additive.. E. Y. hols.7 kcal / mol of stra in to the diaxial form. yf' ~l The measuremen t of the stren gth s of in tramolecular hydrogen bonds can be p erformed by examining the equilibria between differe nt cyclohexa ne ring confor mers." f.-Y.. Such additivity can j) ~ break d own. for cis arrangements. . however. A common example is the 1. Am. the 1. and Anslyn.interaction d is- cussed above. as shown in Figu re 2. Cyclitols by Neutra l Polyaza. The relative stabilities of variou s conform ations can be . the D. ~ " ~ either be cis or trans to each other. = ii For example.H yd rogen Bond ing Receptors: The Strength and Infl uence of Intra molecul ar Hydrogen Bonds Between Vicinal Alco- gen bond as written . The D.11 D . 116.Geq o of the interconversion can be broken down Hu ang. This u sage is distinct from the axial I equatorial terminology. the equilibrium sh ow n to the right involves conformations m issing (i) and conta ining (ii) an intra mo. one can u se the valu es as a tool to measure other energies.3- interaction contributes 3.G«1o was experi menta lly determined using IR Study fo r the Determination of Intramolecular spectroscopy.

If t. t. although dimethylcyclohexane shows only one methyl peak in the 1 we don't often think of it this way. a compound such as trans-1. k = 222 s-1 . eventu- are separated by 100Hz. much higher than for other carbocycles.v under conditions in which the re is no exchange. the ring flipping process is fast on the NMR time scale. This is Note that 2. we expect two signals at tempe ratures signifi- configurations probed in MR spectroscopy are extraor. the rate constant for interconversion slows. we wi ll see one peak. if we lower the tempera- nology di rectly defines the time scale. The result is that we see coalesced peaks (two peaks have become one) in an NMR spectrum H3C~ because of a dynamic process. one expression of which is shown in the equ ation we see two signa ls now depends on the magnitude of the below. and it is summarized in the potential energy diagram given in Figure 2. In fact. Thus. ally reaching a point where two methyl signals can be the units of a first-order rate. but that inter.8 kcal / m ol. Considerin g the range of scale is typically in the ms range.22 • .t must be very sm all. but it is still a useful is in a range that is well match ed to the rates of conforma.3- So. Using the Arrhenius equation we find that at room ple. ventional NMR spectrometers is 5. Although this barrier is high for a cycloalkane. CH 3 . it is important to distinguish between the axial and equatorial positions. in many circum- stances. the molecule enters the twist boat conformer. and the chemical shift separation is about neous. nal at temperatures significantly above room temperature making the NMR time scale relatively slow. like cyclobutane and cyclopentane. time scale (see Table 16. The NMR time suited to the NMR time scale. the rate of intercon version required to achieve coalescence is .12.8 kcal I mol. this equation is not exact. UV spectroscopy involves relatively large energies. it is s till low enough to be very easily traversed at ambi- ent temperatures. If E. If the rate of interconversion is fas t on the NMR time scale (larger than the reciprocal of their peak separation in Hz). (For a dis- cussion of the NMR time scale. Whether ciple.3-dimethylcyclohexane cal shift separa tion of t. common separation. All spectroscopies have an associated titative analysis of such systems. it is 0. Consider again two singlets separated by 100Hz as a The time scale is set by the Heisenberg uncertainty prin. t must be relatively large. and undergoes rapid interconversions among equivalent forms via a boat transition state. For spectroscopy.20 kcal I mol-a very convert through some dynamic p rocess in the molecule. After passing through a half-chair transition state. we are considering the case of two range of activation barriers that ca n be probed with con- nuclei that display distinct NMR signals. convenient range for conformatio nal analysis. The reciprocal of su ch a rate seen. cantly below room tem perature (k << 222 s.1). a process known as pseudorotation. guideline. they are separated by 100 s-1. For more complex because the time scale associated with NMR spectroscopy p rocesses. for a process absorptions can be thought of as essentially instanta. However. the In particular. For two singlets of equal inte nsity that show a chemi. temperature coalescen ce corresponds to E. However. temperatures and chemical shi ft d ifferences accessi ble. and indeed UV and this is a useful number to remember. The boat Going Deeper The NMR Time Scale One of the most p owerfu l tools for investigating confor.106 CHAPTER 2: STRAIN AND STAB ILIT Y As in other cycles. We noted that the barrier to ring-flipping in chair cyclohexane is about 10. lnterconversion of 1. t. Cyclohexane conformational dynamics are well is time.01 s or 10 ms. common NMR termi. what is the NMR time scale? Actually. This conformer is flexible. For exam.) The path by which the chair forms interconvert has been extensively studied. in this case. kcoal = 2. When two peaks ture. Detailed sp ectral fitting programs allow quan- tional processes. the different types of hydrogens (axial and equatorial) interconvert by a conformational process. the energies separating the va rious spin 100Hz. What is unique about cyclohexane is that the barrier for this process is ~ 10. ture. implying that t.::lv mational analysis and a range of other phenome na dis- cussed in this text is dynamic M R spectroscopy. (k > > 222 s.22 is simply 1T I y'2.E is small. see the next Going Deeper highlight. At coalescence. H NMR spectrum.1) and one sig- dinarily small. is 15 kcal I mol. Since this is substan- ti ally less tha n 15 kcal I mol. E is the energy gap between the barrier to the interconversion process and th e tempera- two states that are connected by an absorption . = 15 kcal l mol.1 for a more comprehensive list).

though. Relative energies are given in kcal / mol.BASIC CO NFORMATIO AL A NALY S IS }Q7 Half-chair Flagpole positions \ Boat J M) l 1-2 5.Transannular Effects As soon as w e progress beyond cyclohexane. At no time during the ch air flip process are all six car- bons in one plane.--. Table 2. a structure that can be made more rigid by substitution or ring fu sion. a long range steric interaction called the flag- pole interacti on.(CH2). Figures 2.-- #¥ Twist-boat Chair Chair Figure2.3 10.--._!_. Cyclohexane is also an important component of m any biological structures.12 Conformation al interconversions in cyclohexane. Several conformers are close in energy. as this process has never been directly observed. where n > 6. Note that because of the small barrier. Perhaps surprisingly. ---. common rings (n = 5. The value of cyclohexane as a workhorse of conformational analysis is undeniable.. The larger rings.3 R E L AT IO NS HIP S BE TW EEN S TR UC TUR E AN D EN ERGE T ICS. such as steroids and pyranose forms of sugars (see the next Connections highlight). Eventually.12).15 illustrates this point. cyclohexane has been a platform for innumerable studies of conformation al effects on reactivity. w e return to the more typical mod el of car- bocycles. as this would be a very highly strained structure. among other things. These observations m ake it possible to group cycloalkanes into small rings (n = 3 and 4). 6 and 7). medium rings (n = 8. and large rings (n 2:: 13). With its relatively rigid structure. form is destabilized by.11 C and D emphasize the sh ape of cyclohexane and simple derivatives. Here we describe a few general features of these systems. and they often interconvert over relatively small barriers. one can expect many twist-boat-tw ist-boat interconversions before escape from this local minimum back to the chair manifold. Larger Rings. . The 1-2 kcal / mol barrier for interconversion of twist-boat forms is an estima te. they are not strain free. an important as- pect when con sidering biological recognition processes. the system escapes the twist boat regime via another h alf-chair transition sta te to the "flipped" ch air. The situation rapidly becomes very complicated and a domain primarily for the aficionado of conforma tional analysis. 2 .8 not have large strain energies of the sort seen in the small rings.

A major role of cholesterol is to insert into and tant consequences.2 6. disk sh a pe of chair cyclohex- block is decalin.Decalin trans-Decal in HO''" Cholic acid cis -Decalin trans. m aking these gene rall y The cis ring fusion mus t in volve one axial and one lipophili c molecules adopt an elon ga ted.. This req uirement h as severa l impor. the second ring is joined to the b eca use molecular s hape is always cruci al in biological first by two equ a torial connections.7 9. a nd trans. Most s teroids. one) o r aromati c rings (as in estrone). have all trans fu s io ns ships to the ring fusion. aro und the ring ca n be unambi guously assigned as axial Steroids are importan t fused-ring cyclohexanes. cis ring fu sion.Decalin .6 12.108 CHAPTER 2: S TRA I N AND STA BI L ITY Table 2. Second. In the trans for m. flexib le. cis-decalin is less sta ble than the reby stabili ze cell membranes.9 9 12. but the cis form forces a kink into the system.3 5 6. disk-like struc- equ a to ri al C-C bond. A common building extends the relatively fl at.15 Strain Energies and Group Increment Correction Factors for Cycloalkanes. No chair flipping is p ossible.0 Connections Ring Fusion-Steroids The third key differen ce between cis. (in kcal/mol) Strain Group Strain n energy increment n energy 3 27. The trans form re tains and is by fu sing a second ring onto it. as the ax ial and equ a tori al lin kages ca n intercon. or equatorial based simply o n their cis or tra ns re lation.2 6. the cis for m is ul ar shape is crucial to this fun ction. tra ns ring fusion issue rig id. Th is cis rin g fusion alters the molecula r shape considerably. ca l probes of reacti on mechani sms. This assig nment h as freq uentl y or h ave olefins a t a fu sion (as in cholesterol o r testoster- proven to be a useful s trategy for designin g s tereochemi.6 10 12. and also crea tes an interesti ng juxta- position of the three h yd roxyls.2 11 11 .4 4 26.9 8 9. and no doubt the m olec- lrans-decalin by abo ut 3 kca I I mol.1 0 13 5. the fl a t. Fused a re two ways to fu se two cyclohexa nes.and trans- O ne way to "lock" a cyclohexane into one parti cular chair d ecalin relates to sh ape.4 14 1. Either way. There ane.1 6 0. (CH 2 ) . such as cholic acid (a component of bile) which adopts o ne vert by a chair flippin g process. First.8 16 2. trans-decalin sh ape is maintained. and so substituents is quite important.5 27. tu re.2 7 6.3 12 4.3 26. which combines two cyclohexa nes. e::pH H H / Cis fu sion / " Trans fusion ~ H H HO c:tJ ~ Cholesterol H cis. The re are exceptions. such as cholesterol. The tran s form is qu ite recognition processes.9 15 1. termed cis and cyclohexa nes are common in biological structures. the cis vs.

13 A low energy conformation of cyclodecane shown in ball-and-stick and space-filling (CPK) representations. it is reasonable to assume that a cyclobutane will do the same. the strain in cyclic systems also arises in part from torsional strain. the in- terconversion between ring conformers is predicated on high energy ring torsional motions. all de- rive from vibrational modes. The group increment correction for a cyclobutane is based on fl. 2 .13). a number of interesting reaction mechanisms depend on this effect. Note that these are not identically equal to the accepted strain energies for the parent ring system. a series of correction terms for common ring systems has been developed. Ring Torsional Modes The internal motions that we have discussed for the interconversion of conformers of cyclobutane. Note the close H • • • H transannular contact-a similar interaction occurs on the "back" side of the ring.8 kcal / mol contribution to fl. cyclopentane. For cyclo- . We also saw that a small ring such as cyclobutane leads to a substantial failure for the group increment method. and for n 2:: 17 the rings become so large that they are barely dis- tinguishable from long chains. as discussed. Any strain that primarily arises from non-standard bond angles is called angle strain (or Baeyer strain). because of their proximity. In the group increments method. and those alluded to with the larger rings. this is often called Pitzer strain. Just as we assumed that a gauche butane interaction will make a consistent 0. Again. It is not obvious at first. The strain is often further broken down into small angle and large angle strain. the transannular carbons can react with one another in ways that would not normally be expected.H 1°.Ht values for strained ring systems.H 1° values for a number of structures. What is the origin of the strain in medium-ring compounds? The strain primarily results from a novel feature of these systems that is well illustrated by cyclodecane (Figure 2. There is significant strain in the medium-ring compounds. Representative values are given in Table 2. the strain energy of cyclobutane is specific to the parent compound. Also. and represents an average value that gives the best agreement with the range of experimental data. because of its strain energy. As such. 3 REL ATIO NS HIPS B ETWEEN STRU C TURE A N D E N ERGETIC S. all cyclic alkanes are strained to some extent. In contrast. In the context of rings. This transannular strain cannot be avoided without substantial torsional and I or bond angle distortion.] and then adding appropriate ring strain correction values. with the goal of obtaining accurate fl. The concerted bond rotations lead to the interconverting ring conformers. we assumed that a CH 2 always makes a constant con- tribution to tl. The larger rings show diminish- ing strain with increasing n. With this in mind. by first adding up all the ba- sic group increments [C-(Hh(Ch. some carbons that lie across the ring from one another will end up being fairly close to one another. cyclohexane.15. more than one bond torsion is required (if only one bond rotated the ring would break). it is now possible to predict tl. Just as high energy bond torsions lead to bond rotations. etc. With these new correction terms. medium-sized rings are strained.H1° for a molecule. Group Increment Corrections for Ring Systems Except for cyclohexane and the very large cyclic alkanes. However.H1° values for cyclic systems. although they are quite close. but examination of models shows that in closing up (CH 2 ) 10 into a cycle. With rings.2 A on a side I Figure 2.BA SI C C O NF ORMATION A L AN ALY SIS 109 Triangle is 2.

there is extra strain induced by the ring fusion event. a motion that involves the ring carbons moving toward and away from a half-chair conformation.2]pentane tential surfaces that describe the torsional motions of the atoms in these rings are very shal- low and there is significant motion of the ring atoms.14 tematic nomenclature of more complex multi-ring structures builds on this system. there are m + 1 atoms in the ring system possible. the energy of this motion would be quantized. is very highly strained (Figure 2. and Nomenclature ofbicyclic rapidly becomes unwieldy.O]hexane would be expected to be the sum of the strains of cyclopentane (6. For these systems.2. More typi- cally. Cycloalkenes and Bredt's Rule Incorporating an olefin into a small ring increases the strain of the system. The po- Spiro[2. and these are called the bridgehead car- Bicyclo[2. for both possible ways of combining two cyclopropanes the expected strain energies are 2 X 27. See http: I I www. the strain in bicyclo- [3. the sys- tem is termed spiro.12 ha s substantial barriers. causing the /15 1° for these systems to be more positive than conformationally locked rings. and the nomenclature is fairly straightforward (Figure 2. Therefore. as well as the larger ring systems.4]decane energy diagram in Figure 2.O]butane. When a molecule has two rings and the two share only one carbon in common. We'll avoid it as much as possible here.2 kcall mol) and cyclopropane (27.4. in excellent agreement.l. As much as within th e brackets. but for practical purposes. the two rings have two carbons in common. In an amazing example. As such. The barriers are all near RT.0]nonane Bicyclic Ring Systems Complex.5 kcal l mol). the exception to the norm. roughly 10 kcal l mol more than expected. cyclopropene. as we have already alluded to. Bicyclo(3. and olefins prefer larger angles than alkanes. Thus. it should be more destabilizing to have an olefin in a ring than an acy- clic system.15 shows ring strains for a number of cyclic and polycyclic systems.3 = 158 kcal l mol.14). but these are not all absolutely firm num- for spiro compounds.5 = 55 kcall mol. Here we mention some fundamental issues inherent to multi-ring systems. for example. Is the strain in a bicyclic system simply the sum of the individual ring strains? Perhaps surprisingly. Significant excep tions occur for the smallest rings. which is 33. the strain energy of cubane (Figure 2. mas the sum of the numbers Figure 2. and the experimental value is ~ 166 kcal I mol! It is also true that polycyclic ring systems (if they do not contain cyclopropane and cyclobutane rings) are often very well treated by the molecular mechanics method (see below). if we define .ac. in that the smaller rings enforce compressed C-C-C angles. The effect is consistent with expectation.7kcall mol.15). the barriers to interconversion [><] between conformers are very low. The potential Spiro[5.9 kcall mol. and m + 2 bers. Thus. and with larger systems the additivity approach is usually effective. multi-ring systems figure prominently in many aspects of organic chemistry. The sys- Figure 2. cy- clohexane has a significantly higher energy torsional motion than other rings.3.0]decane. bicyclic compounds. and the concerted torsional motions that lead to the interchange between a chair and twist boat are relatively stiff.ukl iupac I fused ring I for all the gory details. the data are consistent with each other. For example.2]pentane and bicyclo[l.1]heptane bons. However. We will encounter such sys- tems throughout the text.2. for spiro[2.chem systems. this quan- tum mechanical effect ha s little relevance.15) is predicted to be 6 X 26.O]pentane. There will be three different paths between the two common carbons. but the basic patterns can be atoms in the ring system for expected to hold. and little energy needs to be put into the concerted tor- sional vibrational modes to interconvert the conformers. The effect is significantly diminished in bicyclo[2. The actual value is 33. making an analysis of ring strain straightforward. . Once again. Note that.1]octane co Bicyclo(4.l.t- 110 CHAPTER 2: STRAIN AND STABIL ITY butane and cyclopentane. and the bicycle norbornane is named by listing the lengths of these paths. the ring strains are in the 65-66 kcal l mol range. from exotic tests of theory to equally exotic natural products.l. this is very often the case for simple systems. Cyclohexane is. decalin is bicyclo[ 4.qmw. beginning with a bit of nomenclature. These low energy vibrational modes therefore impart favorable entropy to the ring systems. Care must be taken not to overinterpret the precise values.

5 ~ 54.5 A 41 D 26.9 1 ~ ~ ob ~ h lQ & CD 0 I I 24 25 30 34 37 37 38 36 -27 -- CJ ~ ~ Lb & @ 4 ()> cgJ 16.2 0 5.7 0 6.140 -68 """' 31-33 Figure 2. as indicated by the structure on the right. It has long been known that trans- cyclooctene is relatively stable. with a strain energy of . The olefin in this structure is substantially distorted.0 I I i cp cp cp cp H H H H Lb ~ ~ ~ 24-27 13. A related issue is whether it is possible to incorporate an olefin at the bridgehead posi- tion of a bicyclic system.2 H -1 .f--.7 98 104 105 89 -166 . trans-Cycloheptene has been prepared and experimentally characterized at low temperatures. however.0 1 -Bu 1-Bu 11 H 9.16 Two views of a calculated structure for trans-cycloheptene . The view on the left.4 H 15. which is clearly not the case here. Extensive investigations have shown that there are substantial lim- Figure2.. these three will sum to 360°. shows the substantial twisting in the carbon framework. At a planar center.. But what about trans- cyclohexene or trans-cyclopentene? The smallest cycle into which a trans olefin can be embedded has been a subject of considerable study. sighting directly down the C = C bond. . The extent of pyramidalization is indicated by the three valence angles.0 16. Note. as indicated in Figure 2. that the olefin is not only twisted. trans-cyclopropene seems like a pretty unreasonable structure.15 A potpourri of strained molecul es and their associated strain energies (in kcal/ mol).3 D 31-34 D 29 <1> 66. 2 . and it h as an esti- mated strain energy of 27 kcal I mol.16 kcal I mol.3 rn50.BASI C CON FORMATIONAL ANALYSIS 111 6 27. When discu ssing such systems we automatically put the olefin in a cis geom e try-that is.5 [><] 65 ~ II i -··---·-··-.16.9 I 05-6 0> A £b : 33.9 ~ 41 .8 H 1. a significant amount of pyramidalization occurs.2 -50 17. 60M D> 57.3 RELAT IONS HIP S BETW EEN STRU C TURE AND E NE RGET I CS.

In contrast. bio- chemistry. One way to gauge the stability of an olefin is to evaluate its heat of hy- 0 0 drogenation (6. On the other hand. inorganic. and we will discu ss some conformational biases associated with them. Early work suggested 5 = 9 was the limit. A series of bridgehead olefins is shown in Figure 2. strains such as are inherent in anti-Bredt olefins and fused small rings are static. Similarly. and it can be seen that a bridgehead olefin must be trans in one ring of a bicyclic system. where steric clashes and angle constraints imposed by ring systems dominate both structure and energetics. There is variation. A typical example- . cyclohexane is also not strained. our analysis of structure and energetics has been focused primarily on saturated h ydrocarbons and simple olefins. resulting in corrections to the group increments for olefins.c]alkanes. With that as a foundation.Hhyd This 6.15. the stability of the bridgehead olefin follows the pattern for the simpler trans cyclic olefins. These lessons from organic chemistry are applicable to any other field of chemis try-organometallic. but this limit was then dropped to 8 and then to 7. and polym er chemistry. Bredt's rule was an attempt to rationalize which bicyclic systems could tolerate a bridgehead olefin.112 C HAP TER 2: STRAI N AND S TABILITY its to such efforts. Bredt's rule considered bicyclo[a.4 Electronic Effects Thus far. The strains arise from both static and d ynamic processes. olefins that are especially strained because they are at a bridgehead position are still referred to as anti-Bredt olefins. Although the analysis has changed somewhat. 2. Yet ring interconversion motions have higher barriers in cyclohexane because significant strain is introduced during the cou- pled torsions that lead to ring flipping. we will need to invoke orbital mixing arguments to rationalize many observations. Comparable studies of cycloalkynes have been made. the ring strain of cyclopropane cannot be relieved by any dynamic motion. for cyclobutane a small pucker arises that alleviates some of the strain. and this reaction is usually quite exothermic. for linear butane. and strain only arises during a conformational change. some molecules are not strained.1 Interactions Involving TI Systems Substitution on Alkenes We have already seen that there is a steric strain in cis olefins relative to trans. We will fo- cus on both stabilizing and destabilizing interactions involving such groups. Along with steric interactions.b. the reaction of an olefin with H 2 to produce ). Recall from introductory organic chemistry that increased substitution on an alkene leads to a stabilization of one alkene relative to another. it was even- tually realized that better predictions could be made by simply considering the size of the ring that contains the trans olefin. However. Yet. but even cyclopentyne has been trapped as a transient inter- mediate. For example. and is therefore static. Systems in which the olefin can be trans in an eight-membered ring are less s trained than those in a seven-membered ring and so on. and are not relieved by any particular conformational change. and cannot be relieved by a molecular motion. the all staggered form has no strain.24. 2. while others are only partially relieved by a conformational change. There are also stabilizing effects associated with increased substitution of the alkene by alkyl groups. and focused on the minimal value of 5 = a + b + c that allowed a bridgehead olefin. 2. all of which are manifest in weaker bonds. we can now consider more complex systems involving unsaturated groups and heteroatoms. and strain is only introduced during a dynamic process. Summary of Conformational Analysis and Its Connection to Strain Structural distortions give rise to strains within molecules. but the ba- sic trend holds.H is for Eq.4. that of bond rotation. the analogous alkane. The results are similar: cyclooctyne is the smallest isolable system. Similarly. To a good approximation. All these examples show that some strains are in- herent in a structure.



-;- 31


Heat of hydrogenation values (ll.H0 ) for several a lkenes. The effect of alkyl substitution is evident. Also, the
stability of trans double bonds remains relatively constant, but significant des tabilization of the cis alkenes
is seen as th e R group size increases. Derived from data in Turner, R. B., Jarre tt, A. D., Goebel, P., and Mallon,
B. J. " H ea ts of H yd rogenation. IX. Cyclic Acetylenes and Some Miscellaneou s Ole fins." ]. A m. Chern . Soc. 95,

trans-2-butene-is shown, and for most conventional olefins t>.Hhy/ is near the -27.75 kcal /
mol value shown.

trans-2-Butene + H 2 - Butane
-2.58 0 - - -30.37 (Eq. 2.24)

Figure 2.17 shows a collection of heats of hydrogenation for various substituted olefins.
The stabilization due to alkyl substitution is evident. This stabilization is a sizeable effect,
such that cis-2-butene is more stable than propene, despite the adverse steric interaction. The
tetrasubstituted olefin is less stable than the trisubstituted, but less so than expected given
the destabilizing cis interactions. One way to rationalize the stabilization of olefins by alkyl
substitution is an orbital mixing phenomenon of the sort we invoked for propene in Chapter
1. There is some mixing of a filled 1r(CH3 ) orbital with the 1r* orbital, resulting in the LUMO
shown in the margin. This is a filled-empty mixing, and so it must be stabilizing. The contri-
bution of the 1r(CH3 ) orbital here is much smaller than in the HOMO of propene (Figure
LUMO of propene
1.18), in a clear manifestation of the energy gap law. The olefin 1r* MO is higher in energy
than the 1T MO, and so the energy gap to the 1r(CH3 ) orbital is greater and the mixing is less.
Figure 2.17 also shows several cis-trans comparisons. As the alkyl group increases in
size, the energy difference increases. With two t-butyl groups on an olefin, the cis-trans dif-
ference is 10 kcal I mol. Because of these trends, the simple cis alkene group increment of Ta-
ble 2.11 must be used with caution.

Conformations of Substituted Alkenes
The most obvious conformational effect in alkenes is that they are planar. The planarity
is a contra-steric effect that derives from the geometry of the 1T bond . Here, all the groups


around the alkene, including the alkene carbons, are coplanar. However, simple substituted
1T systems have additional conformational features to consider. For example, propene shows
a 2 kcal / mol preference for the eclipsed over the staggered form (Eq. 2.25), with the stag-
gered form being the transition state for rotation.

H\ H''''·· ··'''' H
Eclipsed (preferred) Staggered
(Eq. 2.25)

The origin of the preference for eclipsed over the staggered is subtle. As just discussed
above, in Chapter 1 we showed how a 'TT(CH3 ) orbital will mix with the classical olefin 'TT or-
bital to give the HOMO shown in Figure 1.18. To first order, this interaction is the same for
the eclipsed and staggered forms. However, there is a secondary interaction between C1 and
C3 that can discriminate between the two conformers. This interaction, shown below, can
only occur for the staggered form, and it is necessarily out of phase and is thus destabilizing.
This destabiliza tion of the staggered form has been put forth as the cause of the conforma-
tional preference in propene. In 1-butene, the eclipsed form is still preferred, and surpris-
ingly there is a slight preference for the methyl group to be in the same plane as the alkene.

Secondary orbital interaction in propene

A similar bias is seen in carbonyl compounds. Acetaldehyde shows a 1.0 kcal I mol pref-
erence for the eclipsed form over the staggered (Eq. 2.26), with the latter again being a tran-
sition state for rotation. In propanal the conformation having the methyl group eclipsed
with the carbonyl (Eq. 2.27) is preferred by ~ 1 kcal / mol. Similarly, for 2-butanone and even
Preferred conformation 3-pentanone, the conformations with the methyls eclipsed with the carbonyl are preferred.
of 3-pentanone It is generally accepted that electronic interactions are controlling these preferences, but the
effects are subtle.

Eclipsed (preferred) Staggered
(Eq. 2.26)

H H,,
\ c.;= a (Eq. 2.27)
I ~CH 3

The 1-methylallyl cation displays an interesting conformational effect that is similar to
those discussed above, and is another nice example of how the group orbitals given in Chap-
ter 1 are useful. As shown in Figure 2.18, the structure with the methyl "in" (s-cis) is less sta-
ble than the one with the methyl "out" (s-trans), the energy difference being a surprisingly
large 5-7 kcal/moL One way to rationalize this effect is to examine the HOMO of the ion,
shown in Figure 2.18. This orbital is the out-of-phase combination of a 1r(CH3 ) orbital and the
allyl HOMO. With the methyl tucked in, as in the s-cis form, there is some overlap between


Repulsive interaction Some residual7t bonding

Figure 2.18 0 .

The geometri cal isom ers of 1-methylallyl cation, and the HOMO of the
s-cis form.

7T(CH3 ) and the p orbital at C3. This overlap is n ecessarily out of phase and so is destabilizing ooH-G-C-H torsion angle
to the HOMO and the molecule . Such an interaction is absent ins-trans, thus rationali zing
the conformational preference.
Neighboring group orbita ls also influence the conformations of radical cations. With
ethylene, oxidation of the 'IT bond leads to a structure with a 25° dihedral angle. This dihedral
angle is due to a competition between residual 'IT bonding (preferring a 0° angle) and the abil-
ity of an occupied 7T(CH 2 ) group orbital to mix with and stabilize an adj acent empty p orbital gooH-G-G- H torsion angle
on the other CH 2 g roup (preferring a 90° dihedral angle).

Direct attachment of alkenes, without any intervening atoms (conjugation), leads to 25° H-C-G- H torsion angle
molecular orbitals that are delocalized across all the sp 2 hybridized carbons (see butadiene
'IT as a compromise
in Appendix A). The term conjugated is normally associated with 'IT bonds that are arranged
Ethylene radical cation
in a line or loop, such as with butadiene, cx,l3-unsaturated ketones, or benzene. A cross-
conjugated system is a term used to define conjugation with the 'IT bond arranged in a
branched fashion, such as w ith the structures shown to the side . Either form of conjugation
lead s to delocalization, which s tabilizes the 'IT system. Estimates of the energetic stabiliza tion
due to conjugation vary, depending on the reference structure. Simple analyses, such as
comparing hydrogenation energies of butadiene vs. 1-butene, neglect the stabilizing effect
of the ethyl group in 1-butene and so tend to underestimate the stabilization. Recent esti- Conjugation

mates put the conjugative stabilization at 8 kcal l mol for butadiene and 9 kcal l mol for buta-
diyne. The effects are typically not as large as they are in reactive interm edia tes, where all y lie
stabilization is worth 10-15 kcal I mol.
Cross conjugation
Conjugation favors a planar structure, and for a prototype compound such as 1,3-
butadiene, there are two choices, termed s-trans and s-cis (the s signifies geom etry around a
single bond). Due to sterics, the s-trans conformation is preferre d, and the barrier to inter-
conversion is near 4 kcal I mol. In fact, the s-cis is not actually present at all. It is a transition
state between conformations that are referred to as skew. In the skew conformation, the C1-
C4 repulsion is slightly relieved. This distortion from planarity is further evidence that the
conjugation in molecules such as butadiene is not an energetically strong effect.

(Eq. 2.28)
s-Trans s-Cis

As with conjugated dien es, cx,l3-unsaturated ketones prefer planar conformations. In
pro penal the s-tran s conformation is strongly preferred (- 5 kcal I mol, Eq. 2.29), while in
3-buten-2-one the s-trans is preferred by a factor of about three (Eq. 2.30). This preference is
reversed in 4-methyl-3-penten-2-one (Eq. 2.31), where the s-cis conformation dominates by
about a factor of three (you are asked to rationalize these trends in the Exercises at the end of
the chapter).

00 (Eq. 2.29)


~0 -- ~
(Eq. 2.30)

Ylo -- Y ( (Eg. 2.31)


A Conformational Effect on the Material m agnitude. This example shows that careful control
Properties of Poly(3-Alkylthiophenes) of conformational effects can have dramatic effects on
Polythiophene is a classic conducting organic polymer, m aterial properties.
often u sed to stud y the mechanism of electrical conduc-
s s
ti vity (see Chapter 17 for a discussion of conducting poly-
mers). The mechanism of conduction relies upon effecti ve
conjugation of the 1T system along the polymer and good
-crcr R R

s tacking between adjacent polymer chains. In the stan-
s s s
dard random p olymerization syntheses that start wi th
Polymerize Head-to-head
3-alkylthiophenes as the monomers, one achjeves a vari-
ety of linkages (head-to-tail, head-to-head, and tai l-to-tail, R R
shown to the right). The head-to-head linkage creates a s s
s teric interaction that leads to a conformation along the Tail-to-tail
polymer chain that disrupts the conjuga tion and would
impede stacking between adjacent polymer strands. How- R R
ever, a more directed synthesis can produce a polymer
McCullough, R. D., Tristram-Nag le, S., Willi a m s, S. P., Lo we, R., and jay-
that has pure head-to-tail linkage throughout. This direct a raman, M. "Self-Orienting Head-to-Tail Poly (3-alkylthio phenes): e w
synthesis removes the adverse s teric interactions that Ins ights on Structure-Prope rty Relations hips in Conducting Po lymers."
caused twisting along the backbone, and the conductivity ]. Am. Chem. Soc., 115, 4910-4911 (1993).
of thin films of this polymer is h igher by a few orders of

Many stud ents are undoubtedly familiar with the Hiickel4n + 2 rul e for predicting aro-
m aticity. Any hydrocarbon or heterocycle with 4n + 2 e lectrons in a fully conju ga ted cyclic 'IT
system is considered aromatic. Rem ember, it is the nu mber of electrons, not the number of
atoms that defines a system as aromatic or not. An aromatic system is more stable tha n ex-
pected when compared to similar stru ctures. Benzene is the paradigmatic example of an aro-
m a tic system. As such, several aspects of the chemistry of benzene derivatives are consid-
ered h allmarks of aromatic character-namely, a pattern of substitu tion, not addition, wi th
electrophilic reagents, and an unusu al resistance to oxidation.
An y 4n + 2 system is aromatic, and hence cyclopropenyl ca ti on, cyclopentadienyl
anion, cycloheptatrienyl cation (called tropylium ion), pyrrole, and furan are all aromatic.
They are all unusually stable, as d escribed for cyclopropenyl cation in the following Connec-
tions highlight. Furthermore, they are all planar, like benzene. One would hope for a good
theoretical justification for aromaticity, unifying the similar character of all these com-

0 000
8 0
pounds. Indeed, arom aticity has been studied extensively with electronic structu re theory
m ethods. But, as we discuss in Section 14.5.1, there is still significant d ebate as to the origin
of aroma ticity. We leave a detailed presentation of the theory to tha t section of the book.
Aromaticity is n ot restricted to completely conjugated rings. There are rare cases where
Aromatic species
the geometry of the m olecule allows an orbital overlap such that the compound can be aro-
matic, although there is a saturated center somewhere in the ring. These structures are called


homoaromatic. For example, homotropylium ion has been extensively investigated in this
regard. Furthermore, even trishomocyclopropenium ion is considered to be aromatic, al-
though it also can be viewed as a non-classical carbocation. These structures are also con-
sidered homoconjugated, meaning that conjugation arises between atoms that are not for-
Homotropylium ion
mally a bonded together (see Section 11.5.11 for an example of homoconjugation affecting
In some cases, a resonance structure is required to see an aromatic system. The increased
stability associated with an aromatic system is found for the structure, although the com-
pounds do not appear aromatic unless the resonance structure is considered. Azulene,
which can be drawn as a cyclopentadienyl anion fused to a cycloheptatriene cation, and Trishomocyclopropenium ion
cyclopropenone, which can be written as possessing a cyclopropenyl cation, are two exam-
ples (see margin).
A long standing issue in physical organic chemistry is the quantitative extent to which
benzene is stabilized by aromaticity. Thermochemical strategies have frequently been em-
ployed for such analyses. For benzene the task is relatively straightforward. The standard
analysis is to compare the heat of hydrogenation of cyclohexene to that of benzene. One way
to set up the analysis is to simply calculate tlH 0 for the process of Eq. 2.32. This equation
sh ows what is known as an isodesmic reaction, because the numbers and kinds of Cs and Azulene

H's are equal on both sides of the equation; for example, there are 12 sp 3 and six sp 2 carbons 8
on each side of the equation. Using experimental va lu es for tlHr tlH0 is found to be -35.5
0 0

kcal/ mol, so the reaction is quite exothermic. It is generally accepted that benzene and cy-
clohexane are strain free, but cyclohexene ha s a strain energy of 1.4 kcal / mol, and with this L 11
correction we conclude that the aromaticity of benzene is worth nearly 32 kcal I mol. This Cyclopropenone
thermochemical value agrees well with those determined by other approaches.

3 0 -- 0""' ./.{/
+ 2 0 !!. W = -35.5 kcal/mol
(Eq. 2.32)
!!.Ht -1.2 19.7 -29.4

Antiaromaticity, An Unusual Destabilizing Effect
In contrast to the stability of aromatic systems, a planar '1T system with 4n electrons is
generally observed to be unstable, and is called antiaromatic. The instability of such sys-
tem s is quite severe, as shown with one example in the Connections highlight on the next
page. In Section 14.5.6 we will show that such systems can have biradical character, and


Cyclopropenyl Cation was found that cyclopropenyl cation has an equilibrium
Cyclopropenyl cation is stable enough to be isolated and constant 1013 more favorable than allyl cation-quite an
stored in a bottle, albeit at -20 oc. It can be prepared by as tounding effect!
chloride abstraction from 3-chlorocyclopropene, as shown
OR e
below. The 1H NMR spectrum in nitromethane solvent H
shows a singlet at a very downfield shift: 11.2 ppm.

~I SbCI 5 8 RO~
0 + HOR
+ SbCI 6
Generating cyclopropenyl cation Cyclopropenyl vs. allyl cation

Breslow, R., and G roves, J. T. "Cyclopropenyl Cation. Synthesis and
One can compare the reactivity of cyclopropenyl cat- Characteri za tion ." f. Ant. Clrem. Soc., 92,984 (1970).
ion to allyl cation as a means of estimating the stability
imparted by aromaticity. In the following equilibrium, it


hence are expected to be unusually reactive. The para digmatic example of an antiaromatic
system is cyclobutadiene.
Es timating the destabilization imparted by the antiaromaticity of cyclobutadiene is
more of a ch all enge than estimating the stability imparted to benzene. It was not until late
1999 that D.Ht for this prototype structure was determined by Snyder and Peters. Using pho-
toacoustic calorimetry, a 6.Hf value of 114 ± 11 kcal I mol was determined for cyclobuta-

diene. Using this and other relevant data, we find tha t Eq. 2.33, the direct analogue of Eq.
2.32, is endo thermic by 46 kcal I mol ± 11 kcal l mol. It is not obvious exactly how to incorpo-
rate strain effects, but to first order it is a good approxima tion to say th at Eq. 2.33 is strain
neutral. This analysis lead s to the conclusion that the antiaromaticity of cyclobutadiene is
more substan tial than the aromaticity of benzene. This difference is especially tru e when we
consider these values on a per carbon basis.

20 -- D + 0 t!.W = 45.7 kcal/mol
(Eq. 2.33)
t!. H f
37.5 6.7 114

Cyclobutadiene does not adopt a perfectly square structure. It rapidly interconverts
between two rectangular fo rms. The distorti.on from a square arises from a pseudo-Jahn-
Teller effect, which leads to a lowering of the energy of the system. This dynamic process
(Eq. 2.34) h as a very low activation barrier, estimated to be 5-10 kcal l mol.

D= D (Eq. 2.34)

Cycloocta tetraene (COT) is also antiaromatic when planar. To avoid the antiaromaticity
and severe bond angle strain, the ring puckers into a tub shaped conformation (shown in the
margin) . The barrier to " flipping" the tub is 13.7 kcal l mol. Interestingly, addition of two
electrons or removal of two electrons would create aromatic systems, and indeed both reac-
Cyclooctatetraene ''tub"
tions are known and lead to planar structures.

NMR Chemical Shifts
Aroma ti c systems possess di agnostic 1H NMR chemical shifts. The circulation of elec-
trons wi thin the MOs that are above and below the plane of the ring (ring current) crea tes
magnetic fie lds, giving rise to anisotropy. Anisotropy means that the magnetic field strength
felt by the hydrogens varies as a function of the orientation of neighboring bonds. In an aro-
matic system, the ring current leads to a magnetic field that reinforces the applied fi eld in the
region of the protons, and shifts the proton resonances downfield (hi gher ppm), normally in


Cyclopropenyl Anion
When one deprotonates propene, it is the meth yl hydro-
gens that are the most acidi c. Deprotona tion crea tes the
resonance stabi li zed allyli c a nio n. When the analogou s
reaction is attemp ted with cyclo prope ne, a vinyli c hyd ro-
gen is the one removed . De proton ation of the CH 2 gro up
in cyclopropene (Eq. 2.19) would crea te a n an ti arom a ti c
ani on, an und esirable effect, and thi s reversal in acidities
provided early support for the noti on of destabi liza tion
due to antiaroma ticity.
Schipperijn, A. j . "Chemistry of Cyclopropene. Preparation and Reacti v-
Cyclopropenyl vs . allyl anion
ity of the Cycloprop-1-enyl Anion in Li qu id Ammoni a." Reel. Trn v. Chim.
Pays-Bas, 90, 1110- 1112 (1971).

2 .4 ELE C TRO N I C EFF EC T S 119

the range of 6 to 8 ppm. Antiaromatic systems show upfield shifts, although it is not clear
that this shift is due to a ring current effect. The downfield shift due to aromaticity is defined
as a diamagnetic shift, while a upfield shift is known as a paramagnetic shift.

Polycyclic Aromatic Hydrocarbons
Fused benzene rings also possess the reactivity characteristics
of aromatic systems. For example, naphthalene and anthracene
undergo substitution reactions with electrophiles instead of ad-
dition reactions. Furthermore, using an analysis similar to the iso-
desmic analysis given above for benzene, one can calculate sta-
bilization energies due to aromaticity of 61 and 84 kcal / mol for
naphthalene and anthracene, respectively. Not all polycyclic sys-
tems, however, are simple aromatic molecules. Phenanthrene un-

dergoes addition upon treatment with bromine. Apparently, the
reactivity associated with aromaticity is not as evident in a ring
when the other rings in the compound can be identifi ed as isolated
(non-fused) benzene rings.
Phen anthrene

Large Annulenes
An annulene is a cyclic, fully conjugated hydrocarbon, denoted with the nomenclature
[n]annulene where n is the number of carbon atoms in the ring. Hence, [6]- and [8]annulene
are benzene and cyclooctatetraene, respectively. The aromaticity of these compounds has
been extensively investiga ted. As the ring becomes bigger, there is a lower stabiliza tion im-
parted by conforming to the 4n + 2 rule. As a general rule, it is believed that 22 electrons ap-
proaches the limit of aromaticity, and there is no stabilization for larger aromatic systems.
An example of an intermediate case is [18]annulene. The structure is planar, and its NMR
spectrum shows a strong diamagnetic ring current indicative of aromatic character. As th e
ring gets smaller, an effect analogous to transannular strain arises, which does not allow the
structure to be planar. However, if the compound is bridged, thereby removing the offend-
ing hydrogens, it can become aromatic. An example is 1,6-methano[10]annulene. Further 1,6-Methano [1 O]annulene
discussion of larger annulenes is given in Section 14.5.1.


The stabilization imparted by aromaticity has been
Do not count the circled groups
exploited by nature in the use of the porphyrin ring sys- and the system has 18 electrons-
tem. Shown to the right is the basic porphyrin skeleton. it 's aromatic
If w e exclude two nitrogens and alkenes n ot directly
involved in creating a full y conjugated cycl ic system, we
can see a porphyrin as a (4n + 2) system . Extensive investi-
ga tions have been conducted to explore the exten t of stabi-
lizati on imparted to these structures due to their aromatic
character. It is generally believed that the incorpora tion of
heteroa tom s increases the stabiliza tion, allowing these
large rings to have significantly more stabilization than
the analogous all-carbon ann ulenes.
Franck, B., and Nonn, A. "Novel Porphyrinoid s for Chemistry and Medi-
cine by Biomimetic Synthesis." Angew. Chem . Int. Ed. Eng., 34, 1795 (1995) .


2.4.2 Effects of Multiple Heteroatoms
A basic tenet of organic conformational analysis is that the lessons learned from the exten-
sively studied hydrocarbon systems will carry over more or less unperturbed to systems with heteroat-
oms. The conformational analysis of methyl ethyl ether should not be too different from that
of n-butane. However, in certain cases, when multiple heteroatoms are superimposed on a
h ydrocarbon framework in close proximity, a number of novel "effects" arise, which often
stabilize o therwise unstable conformations. These effects, such as the anomeric effect and
the gauche effect, all have similar origins that can be easily understood using the bonding
models of Chapter 1. Let's start by analyzing bond length effects.

Bond Length Effects
One simple difference in the conformations of pure hydrocarbons and heterocyclic rings
results from the fact that C-heteroatom bonds are of different lengths than C- C bonds.
Bonds to 0 and N are shorter, often causing increased steric strain. Bonds to S are signifi-
cantly longer.

~0 ~N-H ~ 0
Tetrahydropyran Piperidine 1,3-0ioxane 1,3-Dithiane

These differences are reflected in the A values for groups attached to cyclohexane an-
alogues. Table 2.16 shows a number of A values for groups attached to various positions
on heterocycles relative to cyclohexane. Note the dramatic difference between the 2- and
5-positions on 1,3-dioxane. This difference reflects the shorter C-0 bond lengths, making
Dioxane diaxial interaction
the 1,3-diaxial interactions more repulsive.

Table 2.16
A Values (C.G 0 in kcal/mol) for Three Groups in Particular Positions in Heterocyclic Systems
Cyclohexane Tetrahydropyran 1,3-0ioxane 1,3-Dithiane
2-Position 2-Position 5-Position 2-Position 5-Position
Meth y l 1.8 2.9 4.0 0.8 1.8 1.0
i-Pr 2.1 4.2 0.7 1.5 0.8
t-Bu > 4.5 1.4 > 2.7

Orbital Effects
Recall that the introduction of electronegative elements such as F, 0 , and N has a general
effect of lowering the energies of all MOs to w hich they m ake a significant contribution.
Especially important are the low-lying empty MOs (often a * orbitals). In addition, hetero-
atoms introduce lone pair MOs, filled orbitals with very little bonding character that are
relatively high-lying in energy, even though they are associated w ith an electronegative ele-
ment. Many lone pair orbitals also tend to be relatively "localized", presenting a large elec-
tron density at one site for orbital mixing. This combination, high-lying fill ed MOs and low-
lying empty MOs, is perfect for the always stabilizing two center-two electron interaction
discussed in Chapter 1.
It is useful to think of this situ ation as a donor-acceptor interaction. The hi gh-lying
filled orbital donates electrons to the low -lying empty orbital, producing a stabilizing inter-
action. However, this interaction is not electron transfer nor the kind of donor-acceptor in-
teractions often discussed in excited-state phenomena (Section 3.2.4). It is simply orbital


,. -,


Donor orbital
(often a lone pair)
', ', , :
:- Acceptor orbital
(often an antibonding
orbital )

', ,* /
mixing (see above) . No charge-transfer bands are seen in UV / vis spectroscopy and no
highly polarized states are seen . It is still a "cova lent" bonding situation, but there are some
special orbital mixing possibilities.
Within this framework, then, it is useful to classify th e donor and acceptor capabilities of
certain kinds of groups. Useful sequences are shown in the margin. Donors
The trends are fairly standard. Lone pairs are better donors than bonding pairs because Lone pairs> bonding pairs
they are at higher energy. Amongst lone pairs two effects dominate. First, donor ability e:- > N: > 0: (p) > 0: sp2 > F:
increases as electronegativity decreases; and second, donor ability increases as you move I: > Br: > Cl: > F:
C-H > N-H > 0-H > F-H
down a column of the periodic table. These trend s are consistent with the bonding models C-Cl > C-C > C-H > C-F
we developed in Chapter 1. C-I > C-Br > C-Cl > C-F
For acceptor MOs (these are empty u* orbitals), the trends are shown in the margin. C-S > C-C > C-N > C-0
Again, electronegativity and periodic table effects are evid ent. Note that moving down a col-
umn of the periodic table makes a C-X bond both a better donor and a better acceptor. What Acceptors
is dominating here is the polari zability of the X a tom (see Section 1.1.12). Polarizability is not C-F > C-0 > C- N > C- C
explicitly treated in simple orbital mi xing mod els, because it, b y definition, involves the re- C-I > C-Br > C-Cl > C- F
C-S > C- 0
organization of electron density (and hence of orbital shapes) th at occurs in response to an C-P > C-N
interaction. We simply have to treat such effects as an extra layer on top of the simple orbital
mixing models.
Given this information, we would predict that molecules would adopt conformations
that maximize interactions between good donor orbitals and good acceptor orbitals (i.e.,
that maximize the mixing of filled with empty orbitals). The only remaining issue is the pre-
ferred geometry of the interacting orbitals, which is a bit counterintuitive. As shown in Fig-
ure 2.19, the optimal arrangement places the donor orbital anti to the C-X bond that is acting as the
acceptor. The rea son for thi s arrangement is th e unique nodal character of au* orbital. Figure
2.19 shows the interaction of a generic lone pair donor with au* orbital, the latter modeled


Figure 2.19
Preferred geometry for the interaction of a donor (shown as a lone pair)
w ith an acceptor u* orbital. The u* orbi tal is modeled after the LUMO of
CH3Cl, shown at the bottom.


after the LUMO of CH3 Cl. There is considerable cr* orbital density on the backside of the car-
bon, and the interaction of this density with the lone pair orbital is more extensive in the anti
arrangement. Also, while it is difficult to portray in the picture, detailed orbital analyses re-
vea l that the syn orientation experiences both favorable and unfavorable interactions. The
net effect is that the anti arrangement is preferred . We are now ready to predict molecular
We begin w ith a simple system that very nicely illustrates the key principles. Consider
(fluoromethyl)amine, FCH 2NH 2 . As shown in Eq. 2.35, thi s system is perfectly set up for a
donor-acceptor interaction. The preferred conformation puts the nitrogen lone pair (donor)
anti to the C-F bond, optimizing the donor-acceptor interaction . This is really an optimal
case, and the conformational preference is substantial.


(Eq. 2.35)

Another simple system is 1,2-difluoroethane, in which the conformation with the fluo-
rines gauche is preferred over the anti by 1.8 kcal / mol, in what would appear to be contrary
to conventional steric arguments (Eq. 2.36). Donor-acceptor analysis explains the result.
Aligning the fluorines anti places a poor donor (C-F bond) anti to a good acceptor (C-F
bond). However, having the fluorines gauche places the two good acceptor bonds (C-F) anti
to C-H bonds (Eq. 2.37). While C-H bonds are not especially strong donors, they are better
donors than C-F bonds, and so the gauche conformation is preferred. In this case a favorable
orbital interaction stabilizes what would otherwise be a strained structure.

H ,,
1H _ \ . _ t ,,H (Eq. 2.36)
HW - H"'~-
F~ H - H~H (Eg. 2.37)

A classic examp le of donor-a cceptor interactions is seen in hydrogen peroxide, which
also introduces an additional effect that can arise when several polar bonds are present in a
molecule. In H 20 21 solely steric arguments predict that the preferred conformation should
have an H-0-0-H dihedral angle of 180°. This steric argument is augmented by a second ef-
fect. In the anti conforma tion the two large 0-H bond dipoles are aligned anti to one another,
often a significantly stabilizing effect. However, the anti conformer is opposed by donor-
acceptor effects (see structures below).
The 0-H bond is an excellent acceptor, and the best donor is an 0 lone pair. We know
tha t such an oxygen ha s two types of lone pairs, a cr(out)-type orbital that is roughly an sp2
hybrid and a pure p orbital (see Section 1.3.3, and water in Appendix 3). The p-type lone pair
is hig her in energy, and so by the energy gap law we expect it to mix with the accep tor orbital.
As sh own below, this mixing would favor a 90° dihedral angle. The final geometry reflects a
com promise among the various interactions, producing a dihedra I angle of - 120°. A similar
effect might be expected for S-S bonds, and as the following Connections highli ght show s,
the preferred angle is 90°.


Steric preference Donor-acceptor Compromise



Protein Disulfide Linkages are now diastereomeric rather than enantiomeric (see
The conformational preferences of dia lkyl disulfides are Chapter 6 for defi nitions). Thus, any protei n that has a
similar to those of hydrogen peroxide. The dihedral angle single disulfide can exist in two diastereomeric forms,
is - 90° in a typical molecule such as dimethyl disulfide, differing in the geometry around the C-S-S-C bond.
perhaps because the dipole effect is smaller (Sis less elec- If there are n disulfides, 2" diastereomers are expected
tronegati ve than 0). Disulfides are common components (assuming there is no global symmetry in the protein).
of protein structures, formed by linking the sidechains of
the amino acid cysteine. Invariably, su ch disulfide link-
ages in proteins are approximately gauche. Just like
gauche butane, a gauche disulfide is chiral, and so exists
s- s
~-.s-s' R

The two enantiomeric forms
in two enantiomeric forms. ln the context of a protein, of a simple disulfdide
which is always chiral, the two disulfide gauche forms

A particularly important conformational phenomenon that can be explained using the
types of arguments developed here is the anomeric effect of carbohydrate chemistry. The
an om eric effect can be defined as a contrasteric bias toward the axial (a) glycosidic linkage at
the acetal carbon over the equatorial (!3). This preference results from aligning the exocydic
C-0 bond anti to a lone pair of the oxygen in the ring (see below). Such conforma tiona! pref-
erences are extremely important in carbohydrate chemistry, so much so that the central car-
bon involved (Cl of a sugar) is often referred to as the anomeric carbon. Since the formation
of such acetals and ketals is generally reversible, it is a simple matter to equilibrate axial and
equatorial groups at anomeric centers and directly determine which is the more stable form
without resorting to calorimetry or other more complicated procedures.

, polarization

o0 0R

Anomeric effect

Figure 2.20 shows four examples of the anomeric effect. In each case, the large group on
the an om eric carbon of the pyranoside prefers the axial posi tion. The m agnitude of the pref-
erence depends upon the group and the substituents on the ring. However, it is also influ-
en ced by the polarity of the solvent. For the third entry in Figure 2.20, the axial preference is
larger in carbon tetrachloride than in acetonitrile. One might have expected the donor-
acceptor interaction to be enhanced in the more polar solvent due to the polarization im-
plied by the effect (see above). However, the opposite is found. The preferred conformation
of the spirocycle shown as the last entry of Figure 2.20 is the one w h ere each C- 0 bond is
antiperiplanar to an oxygen lone pair orbital.
The simple model developed here provides a convenient way to explain and predict

variations in structure seen in more complex systems. There is, however, some controversy
concerning the anomeric effect. While most chemists accept that a donor-acceptor interac-
tion of the kind show n above exists, there are clearly other factors. For example, an axial ar-
rangement of the exocyclic C-0 bond cancels dipoles, a potentially favorable effect. Such an OR
effect is expected to be most important in low polarity solvents, perhaps explaining the sol- Dipole cancellation
vent effect shown as the third entry of Figure 2.20.
Because of their common physical origin, the various donor-acceptor effects discussed
here have been collectively called the gauche effect. The best conformation of a molecule
has the maximum number of gauche interactions between adjacent lone pairs and / or polar


~~~ 1.5

~CI ~ Cl

~OCH3 9 OCH3
Keq =3.4 in CCI 4
Keq =1.8 in CH 3 CN

Both oxygens axial
is the preferred conformation

Figure 2.20
Top: Three glycosidic systems for which the large group prefers the ax ial position. Bonner, W. A.
"The Acid-Catalyzed Anomerizati on of the o-Glucose Penta Acetates. A Kinetic Thermodynamic and
Mech anistic Stud y." f. Am. Chem. Soc. 73,2659 (1951). Anderson, C. B., and Sepp, D. T. "Conformation and
the Anomeric Effect in 2-H alotetrahydropy rans."j. Org. Chem. 32, 607 (1967). Elie l, E. L., and Giza, C. A.
"Conformational Analysis. XVIII. 2-Alkoxy- and 2-Alkylthiotetrahydropyrans and 2-Alkoxy-1,3-
Dioxanes. Anomeric Effect." f. Org. Chem. 33, 3754 (1968). Bottom: Another example of a strong
conformational bias introd uced by the a nomeric effect.

bonds. Thus, in FCH2 CH2F the polar bonds are gauche, and in a peroxide or a hydrazine
(R2 NNR2) the lone pairs are gauche. It is called the gau che effect, but its origin is the prefer-
ence for having lone pairs anti to acceptor a * orbitals, rather than anti to one another.
An argument based upon an analysis of the rela tive orienta tion (stereochemistry) of or-
bitals is called a stereoelectronic effect. The placement of a lone pair orbital antiperiplanar
to a polarized acceptor bond is jus t our first example. We will see stereoelectronic effects on
reactivity in several places in this book.
Now that we have exa mined organic conformational analysis and various effects that
lead to strain and s tability, let's look at some structures where chemists have put these no-
tions to the test.

2.5 Highly-Strained Molecules

One of the fundamental goals of physical organic chemistry has been to establish the limits
of our models for structure and energetics. How long can a C-C bond be? How much angle
strain can a molecule tolerate? How crowded can a structure be? Such questions h ave de-
fined m any brilliant research efforts and have produced a fantastic array of bizarre and won-
derful structures. Here we present a collection of representative highly-strained m olecu les,
with an emphasis on the structural concepts that are being tested.

2.5.1 Long Bonds and Large Angles
Typical C-C bond len~ths were noted in Chapter 1, and while there is considerable vari-
ation, a C-C bond :5: 1.59 A is not considered exceptional. Many compounds with bonds >


1.611 A 1.64A 1.67 A
n- Bu Ar
1-Bu /
1-Bu ""/C-H

1 1 Ph
c - c ..
(() 'n-Bu

1 / Ar
c - c ..

Related Interests