The Chemistry of Peroxides - Rilegato

Rilegato

ISBN 10: 0470862742 ISBN 13: 9780470862742

Casa editrice: John Wiley & Sons Inc, 2006

Riassunto

The Chemistry of Peroxides is a new volume in the Chemistry of Functional Groups series. This series covers all aspects of organic chemistry with each volume containing chapters on:

• General and theoretical aspects
• Computational approaches
• Thermodynamics and kinetics
• NMR and ESR
• Mass Spectrometry
• Spectroscopies
• Analytical aspects
• Reaction mechanisms
• Syntheses
• Biological effects
• Environmental effects
• Industrial applications

Edited by Zvi Rappoport, this series provides outstanding reviews on all aspects of functional groups in analytical, physical, synthetic and applied chemistry.

Su questo libro

Informazioni sull'autore

Professor Zvi Rappoport, Department of Organic Chemistry, The Hebrew University of Jerusalem, Jerusalem, Israel.

Dalla quarta di copertina

Patai Series: The Chemistry of Functional Groups

Series Editor: Zvi Rappoport

Volume Editor: Zvi Rappoport

The chemistry of Peroxides, Volume 2

Part 1

Part 2

Estratto. © Ristampato con autorizzazione. Tutti i diritti riservati.

The Chemistry of Peroxides, Volume 2, Parts 1&2 (2V set)

John Wiley & Sons

Copyright © 2006 John Wiley & Sons, Ltd
All right reserved.
ISBN: 978-0-470-86274-2

Chapter One

General and theoretical aspects of the peroxide group

ROBERT D. BACH

Department of Chemistry and Biochemistry, University of Delaware, Newark, Delaware 19716, USA Fax: +1 302 831 6335; e-mail: rbach@udel.edu

I. INTRODUCTION 2 II. BACKGROUND AND HISTORICAL PERSPECTIVE 3 III. BOND DISSOCIATION ENERGIES OF SELECTED PEROXO COMPOUNDS 5 IV. THE CHEMISTRY OF PEROXYNITROUS ACID 7 A. Historical Overview 7 B. Peroxynitrite Anion and Peroxynitrous Acid. Ground State Properties 8 C. Higher-lying Metastable States of Peroxynitrous Acid 10 D. Dissociative Pathways for Peroxynitrous Acid 13 E. HO-ONO Reactivity 14 1. Two-electron oxygen atom transfer to N, S, P and Se nucleophiles 14 2. Epoxidation of ethylene and propylene 17 3. Peroxynitrite anion oxidations 21 4. A comparison of peroxynitric acid and peroxynitrous acid 21 5. One-electron oxidation. The oxidation of methane with metastable peroxynitrous acid 22 6. The 1,2-rearrangement of peroxynitrous acid to nitric acid 25 V. THE CHEMISTRY OF DIOXIRANES 26 A. Background 26 B. Molecular Orbital Treatments of Dioxirane 27 C. The Electronic Structure of Dioxirane and Carbonyl Oxide 29 D. Oxygen Atom Transfer from Dioxiranes and Carbonyl Oxides 32 1. A comparison of DFT theory with higher-level methods 34 2. Epoxidation of alkenes with carbonyl oxides and dioxiranes 35 3. Relative rates of dioxirane epoxidation in solution 40 4. Oxidation of saturated hydrocarbons 44 VI. THE EPOXIDATION OF ALKENES WITH PERACIDS 48 A. Early Mechanistic Studies 48 B. Hartree-Fock Level Theoretical Calculations 48 C. Synchronous versus Asynchronous Transition States 50 D. Gas-phase Epoxidation of Selected Alkenes 58 E. Factors Influencing the Rate of Epoxidation 58 F. Epoxidation of Allylic Alcohols 65 VII. OXYGEN ATOM TRANSFER FROM SELECTED HYDROPEROXIDES 67 A. The Oxidation of Amines 67 B. The Oxidation of N, S, P and Se Containing Nucleophiles 70 C. Electronic Factors Influencing Oxygen Atom versus Hydroxyl Transfer 72 D. Model Studies on Enzymatic Oxidation of Heterocycles and Aromatic Rings 77 VIII. MISCELLANEOUS OXYGEN TRANSFER REACTIONS 82 IX. ACKNOWLEDGMENTS 84 X. REFERENCES AND NOTES 85

I. INTRODUCTION

Since dioxygen ([O.sub.2]) is the second most abundant molecule in the atmosphere, it should not be too surprising that O-O bonds play such a major role in our lives. In a great many instances dioxygen is the source of the oxygen atoms used in the formation of peroxo compounds. The peroxo linkage is vital to both the incorporation of oxygen in the human body by biochemical syntheses and biochemical decomposition of molecules in our metabolism. It also is a major player in oxidative degradation, combustion, atmospheric and stratospheric chemistry, as well as in smog reactions. Thus, both the formation and decomposition of compounds containing the O-O bond pervade our lives in many ways and a thorough understanding of the mechanistic nuances of such chemical processes are of vital importance to us all.

More than two decades have passed since the last critical review in this Patai series was presented by Cremer. This comprehensive review considered mainly the conformational aspects, physical properties, molecular orbitals of relatively small peroxo compounds such as X[O.sub.2], [X.sub.2][O.sub.2], peroxides, peroxyacids and ozonides (X = H, C, N, O and F). At that time, the peroxo systems were considered to be among the most difficult functional groups for computational treatment owing to the very nature of the oxygen-oxygen bond. In fact, early theoretical studies on the conformational properties of hydrogen peroxide presented difficulties in just computing the O-O rotational barrier in HO-OH; processes involving O-O bond cleavage were treated with great caution. Many of these difficulties were simply a manifestation of the level of theory available at that time and the speed and sophistication of the computers available. The most commonly available method of calculation was Hartree-Fock theory (HF) and it soon became obvious that such single reference methods were not adequate for calculations involving the numbers of electron lone-pairs inherent to the O-O bond.

In the past decade more efficient code and faster computers have allowed the use of electron-correlated methods of calculation and this has opened this area of theoretical chemistry to a wide range of research groups. The present sequel will be focused mainly on recent theoretical studies on a variety of oxidative processes involving oxygen atom transfer. We will include an extensive description of the very recent chemistry of peroxynitrous acid (HO-ONO), dioxiranes, peracids and alkyl hydroperoxides. Since chemically realistic molecular systems can now be treated at an adequate level of electron-correlated theory, new and exciting mechanistic details of oxidation chemistry can be gleaned from such computations that simply were not available to earlier investigators.

II. BACKGROUND AND HISTORICAL PERSPECTIVE

During the formative years of the development of theoretical procedures for the treatment of oxidative processes involving the peroxo moiety, the Hartree-Fock level of theory was most often applied. There were essentially no serious efforts to treat the O-O bond at the extended Huckel level or by the semiempirical CNDO/2 or INDO methods. In fact, most such semiempirical methods are not specifically parameterized even today for the O-O bond. When the earlier versions of the Gaussian suite of programs (G-70) were widely distributed among academic institutions, the use of ab initio calculations for mechanistic studies became widespread although most researchers were still restricted to the use of a minimal basis set (STO-3G). This 1970 version had basis sets with s and p functions only and no gradient optimization methods or electron-correlation corrections were available. As Gaussian 80 and G82 became even more widely distributed, second-order Mller-Plesset perturbation theory (MP2) with the ability to handle d functions came into limited use with SCF gradients and MP2 first derivatives. Although the Gaussian suite of programs was widely used, other codes such as Gamess, ACES, MOLPRO, CADPAC, Jaguar etc. were added to the arsenal of the theoretical chemist. In fact, many such programs became sufficiently 'user friendly' that a number of experimentalists became adept at doing theoretical calculations to augment their laboratory experiments.

As Gaussian 92 became more widely available to the general research population, standard protocol was to use MP2 or higher-order Mller-Plesset theory up to MP4 to calculate the electron-correlation correction. However, the computational expense involved still typically mandated that the geometry be optimized at the HF level. While the use of HF theory proved satisfactory in some smaller systems, it also often led to major errors in the overall energetics of reactions, especially where lone-pair electrons were involved. The size of the practical basis set had expanded to 3-21G and many applications employed the 6-31G basis set and some even with d-functions on all heavy atoms [6-31G or 6-31G(d)]. Minimal basis sets such as STO-3G were no longer publishable by the late 1970s while 4-31G basis sets were still acceptable for publication in major journals up to the late 1980s. By the mid 1980s the 6-31G(d) basis set had become the standard for most applications unless the size of the system was prohibitive.

The G92 program marked the introduction of density functional theory (DFT) although general skepticism prevented its more general use until the G94 code became available. During the early 1990s it was possible to optimize the geometry of some systems using the electron-correlated MP2 method and this recipe coupled with an MP4 energy correction became the most generally accepted method to study systems up to eight heavy atoms (nonhydrogen atoms). This general protocol was accepted by all but the most rigorous theoreticians to be at least adequate. However, this was not a panacea since calculations of transition structures involving the problematic O-O bond and related molecules sometimes led to structural problems even with the MP2 method. This problem was exacerbated by the fact that analytical second derivative calculations were not yet available, so the optimized structure could not be characterized by a frequency calculation as a minimum (all real frequencies) or a first-order saddle point (a transition state with one imaginary frequency).

In the mid-nineties more highly correlated methods such as CCSD and QCISD became available through distribution of the ACES and Gaussian 94 programs. Geometry optimization with these more cpu intensive programs was restricted for the most part to six heavy atoms. About this time multiconfigurational self-consistent-field (MCSCF) or complete active space (CASSCF) methods became the rage. Such calculations were highly touted as being very accurate and especially good for fairly small biradicaloid systems where more than one reference state was anticipated. However, while this may be true for highly symmetrical alkenes and dienes, the choice of the active space actually used in more complex systems is highly subjective and can lead to serious problems. A cautionary note typically accompanies the suggested use of these multireference methods; these are not the 'black box' calculations so typically available today. During these earlier years most practitioners preferred to use the Gamess code for MCSCF calculations since it was faster than Gaussian. However, a major drawback of these CAS methods existed in that a second-level electron-correlation correction to the total energies was essential in order to be able to compute relative energies of saddle points on a reaction surface.

The implementation of Gaussian 98 and the introduction of much faster computers coincided with the rise of the G1, G2 and CBS-Q methods, another milestone in computational chemistry, since chemical accuracy was now available for compounds up to six heavy atoms. Currently, the G3 and CBS-Q methods can treat systems up to ten heavy atoms without too much difficulty, affording bond energies in most cases within 1-2 kcal [mol.sup.-1] accuracy.

It was not until the introduction and widespread use of density functional theory (DFT) that reliable calculations on larger molecules became possible. This method is much faster than MP2, implicitly corrects for at least part of the electron correlation and also provides geometries and overall energetics in many cases comparable to those of higher-level methods. The B3LYP variant of the DFT code has proven to be especially tractable for systems as large as fifty heavy atoms even with a respectable size basis set. In many applications now being reported a 6-311+G(d,p) or 6-311+G basis set is applied to the more difficult problems. The plus basis (+) is especially useful for the peroxo moiety since its function is to better describe anions or electron lone-pairs through the introduction of a large p-orbital on each heavy atom. In addition to the d-orbitals on each heavy atom (d), polarization functions (p) are included on each hydrogen atom to provide a better description of secondary- and hydrogen-bonding interactions. Today the B3LYP variant of DFT calculations is the method of choice for most investigators working on practical theoretical problems of oxidative chemistry. Direct comparison of B3LYP data with that of other methods by a number of investigators has proven its general applicability.

The scientific community has also been particularly fortunate that the software required to locate such complex minima and transition structures has kept pace with the accompanying explosion in hardware. It is now common place for individual investigators to have in their own laboratory a multiprocessor computer of equal computing power to the regional supercomputer laboratories of just a decade ago. However, without the more advanced code for locating such complex molecules we could not take advantage of such hardware developments. A decade ago most geometry optimization methods used the algorithm developed by Schlegel and introduced in the early Gaussian versions as the 'Berny algorithm'. Today, with such added options as modredundant optimization (redundant coordinate optimization) the number of gradient cycles required to locate the minimum-energy geometry of a saddle point is more than cut in one-half. The author of this review, an experimentalist, published his first theoretical paper on electrophilic addition to alkenes using extended Huckel theory in 1970. He has had the good fortune to watch the evolution of theoretical chemistry to the point where one can now do a systematic theoretical study of alkene epoxidation with peracids and dioxiranes on chemically relevant molecules. At present, state-of-the-art calculations modeling enzymatic reactions are feasible. It is from this backdrop that we now present the more successful theoretical studies on several new and emerging areas of oxidative chemistry involving the O-O bond.

III. BOND DISSOCIATION ENERGIES OF SELECTED PEROXO COMPOUNDS

The chemistry of peroxides is to a first approximation simply dictated by the fact that the bond energy of the generic O-O bond is quite low. Consequently, many reactions of the peroxo moiety are thermally induced since it is assumed that O-O bond cleavage can be accomplished at relatively low temperatures. This is something of an enigma since homolysis produces alkoxyl radicals (RO) that are not especially stable. However, the lack of thermodynamic stability of most peroxides is also at the root of a major problem; difficulty in isolation and characterization of a great many peroxo compounds means that the accuracy of O-O bond dissociation energies (BDE) is sometimes questionable. Molecules possessing an O-O bond, which is electronically challenged with four pairs of lone-pair electrons, pose an interesting theoretical puzzle. Just what is responsible for the relatively weak O-O bond? Since by nature the oxygen radicals that result from O-O bond cleavage are less stable than many typical radicals, including carbon radicals, one could anticipate a much stronger O-O bond.

(Continues...)

Excerpted from The Chemistry of Peroxides, Volume 2, Parts 1&2 (2V set) Copyright © 2006 by John Wiley & Sons, Ltd. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.

Dettagli dell'edizione

• Editore: John Wiley & Sons Inc
• Data di pubblicazione: 2006
• Lingua: Inglese
• ISBN 10: 0470862742
• ISBN 13: 9780470862742
• Rilegatura: Copertina rigida
• Numero di pagine: 1100
• Redattore: Rappoport Zvi
• Contatto del produttore: non disponibile
• Persona responsabile: non disponibile