ORGANIC REACTION MECHANISMS 1987...Ring-chain tautomerism in 1,3-oxazines (10) shows equilibrium constants which may be correlated with substituents in the aromatic residue by a Hammett

ORGANIC REACTION MECHANISMS 1987

An annual survey covering the literature dated December 1986 to November 1987

Edited by

A. C. Knipe and W. E. Watts University of Ulster, Northern Ireland

An Interscience@ Publication

JOHN WILEY & SONS Chichester - New York - Brisbane - Toronto - Singapore

ORGANIC REACTION MECHANISMS * 1987

ORGANIC REACTION MECHANISMS 1987

An annual survey covering the literature dated December 1986 to November 1987

Edited by

A. C. Knipe and W. E. Watts University of Ulster, Northern Ireland

An Interscience@ Publication

JOHN WILEY & SONS Chichester - New York - Brisbane - Toronto - Singapore

Copyright 0 1989 by John Wiley & Sons Ltd.

All rights reserved.

No part of this book may be reproduced by any means, or transmitted, or translated into a machine language without the written permission of the publisher.

Library of Congress Catalog Card Number 66-23143

British Library Cataloguing in Publication Data:

Organic reaction mechanisms. I. Organic compounds. Chemical reactions. Mechanisms-Serials 547.13’9

ISBN 0 471 92078 9

Phototypeset by Thomson Press (India) Ltd. Printed and bound in Great Britain by the Anchor Press Ltd, Tiptrec, Esscx

Contributors

R. A. AITKEN

R. A. COX

M. R. CRAMPTON

G. W. J. FLEET

A. FRY

P. HANSON

A. C. KNIPE

R. B. MOODIE

R. A. MORE O’FERRALL

A. W. MURRAY

D. C. NONHEBEL

M. I. PAGE

R. M. PATON

J. SHORTER

W. J. SPILLANE

Department of Chemistry, University of St. Andrews, Purdie Building, St. Andrews, Fife KY16 9ST, Scotland

Department of Chemistry, University of Toronto, 80 George Street, Toronto, Ontario M5S 1A1, Canada

Department of Chemistry, Durham Uni- versity, Durham DH1 3LE, UK

Dyson Perrins Laboratory, Oxford Uni- versity, South Parks Road, Oxford OX1 3QT, UK

Department of Chemistry and Biochem- istry, University of Arkansas, Fayette- ville, Arkansas 72701, USA

Department of Chemistry, University of York, Heslington, York YO1 5DD, UK

Department of Chemistry, University of Ulster at Coleraine, Coleraine, Co. Londonderry BT52 lSA, Northern Ireland

Department of Chemistry, The University, Exeter EX4 4QD, UK

Department of Chemistry, University College, Belfield, Dublin 4, Ireland

Department of Chemistry, The University, Dundee DD1 4HN, Scotland

Department of Pure and Applied Chem- istry, University of Strathclyde, Thomas Graham Building, Glasgow G1 lXL, Scotland

Department of Chemical Sciences, The Polytechnic, Queensgate, Huddersfield, West Yorkshire HD1 3DH, UK

Department of Chemistry, Edinburgh University, West Mains Road, Edin- burgh EH9 3JJ, Scotland

Department of Chemistry, The University, Hull HU6 7RX, UK

Department of Chemistry, University College, Galway, Ireland

Preface

The present volume, the twenty-third in the series, surveys research on organic reaction mechanisms described in the literature dated December 1986 to November 1987. In order to limit the size of the volume, we must necessarily exclude or restrict overlap with other publications which review specialist areas (e.g. photochemical reactions, biosynthesis, electrochemistry, organometallic chemistry, surface chemistry and heterogeneous catalysis). In order to mini- mize duplication, while ensuring a comprehensive coverage, the editors conduct a survey of all relevant literature and allocate publications to appro- priate chapters. While a particular reference may be allocated to more than one chapter, we do assume that readers will be aware of the alternative chapters to which a border-line topic of interest may have been preferentially assigned.

One change of author has taken place since last year and we particularly welcome Dr. Derek Nonhebel on his return to the fold. He replaces Dr. David Cowley who has made a major contribution to the series over the past nine years by contributing authoritative reviews of Radical Reactions, initially as a co-author with Dr. Nonhebel.

Once again we wish to thank thc publication and production staff of John Wiley & Sons and our team of experienced contributors for their efforts to ensure that the standards of this series are sustained. We are also indebted to Dr. N. Cully, who compiled the subject index.

A.C.K. W.E.W.

Contents

1 .

2 . 3 . 4 . 5 . 6 . 7 . 8 . 9 .

10 . 11 . 12 . 13 . 14 . 15 .

Reactions of Aldehydes and Ketones and their Derivatives by

Reactions of Acids and their Derivatives by W . J . Spillane ............... Radical Reactions: Part 1 by P . Hanson ...................................... Radical Reactions: Part 2 by D . C . Nonhebel ............................... Oxidation and Reduction by G . W . J . Fleet .................................. Carbenes and Nitrenes by R . A . Aitken ...................................... Nucleophilic Aromatic Substitution by M . R . Crampton .................. Electrophilic Aromatic Substitution by R . B . Moodie .....................

Nucleophilic Aliphatic Substitution by J . Shorter ........................... Carbanions and Electrophilic Aliphatic Substitution by A . C . Knipe Elimination Reactions by R . A . More O’Ferrall ............................ Addition Reactions: Polar Addition by A . Fry ............................... Addition Reactions: Cycloaddition by R . M . Paton ........................ Molecular Rearrangements by A . W . Murray ...............................

M . I . Page ...........................................................................

Carbocations by R . A . Cox ......................................................

1 29 91

139 197 247 271 287 299 315 345 369 387 419 457

Author Index. 1987 ....................................................................... 575 Subject Index. 1987 ....................................................................... 629

Organic Reaction Mechanisms 1987 Edited by A. C. Knipe and W. E. Watts 8 1989 John Wiley & Sons Ltd.

CHAPTER 1

Reactions of Aldehydes and Ketones and their Derivatives

M. I. PAGE

Department of Chemical and Physical Sciences, Huddersfield Polytechnic

Formation and Reactions of Acetals, Ketals, and Orthoesters . . . . . . . . . . . . .

Compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Non-enzymic Reactions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Enzymic Reactions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

Reactions and Formation of Nitrogen Derivatives, Schiff Bases, Hydrazones, Oximes, and Related Species. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 C-C Bond Formation and Fission; Aldol and Related Reactions. . . . . . . . . . . 1 1 Other Addition Reactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Enolization and Related Reactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Hydrolysis and Reactions of Enol Ethers and Related Compounds. . . . . . . . . . . 20 Other Reactions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

1 Hydrolysis and Formation of Glucosides, Nucleosides, Oxazines, and Related

Formation and Reactions of Acetals, Ketals, and Orthoesters

Weaker than usual anomeric effects are observed in 2,3- and 2,5-t-butoxy and -trimethylsiloxy derivatives of 1,4-dioxane. Calculations indicate that this is not a steric but an electronic effect resulting from electron donation by the substituent.'

An analysis of X-ray data ofCOCOC fragments shows that there is no correlation between CO bond lengths or COC bond angle and torsion angles whilst the bond angle OCO at the anomeric centre is very sensitive to conformational variation2

The pH-rate profile for the hydrolysis of the cyclic acetal of trans-t,3- cyclohexanediol (1) shows seven inflections, only one of which is due to the ionization of the dimethylamino group. The other inflections are attributed to changes in rate-limiting step and mechanism. Because of strain in the acetal, ring- opening is rapid and reversible. At pH > 7 the rate-limiting step is water or hydroxide ion attack on the oxocarbocation intermediate but at higher pH it becomes pH-independent r ing-~pening.~

There have been further reports on the general acid-catalysed and uncatalysed hydrolysis of acetaldehyde methyl phenyl a ~ e t a l s . ~

The acid-catalysed reaction of alkyl b-tetrahydropyranyl acetals containing an enamine (2) form products in dry methanol consistent with exo- and endo-cyclic C-C bond ~ leavage .~

1

2 Oryunic Reuction Meclzunisms 1987

Acetals of 2-hydroxybenzaldehyde may be prepared from acylatcd salicylaldehyde under basic conditions such as sodium methoxide in methanol. This is attributed to hemiacetal ion formation and subsequent intramolecular acyl transfer (3). Loss of acetatc from the acylal would generate a highly resonance-stabilized intermediate (4) which would be trapped by alcohol to give the acetal product.6

(5) (6) (7)

Divalent metal ions are very effective catalysts for the hydrolysis of acetals containing suitably placed pyridine and carboxylate residues, even though mctal- ion binding to the acetals is weak. Catalysis is due to metal-ion stabilization of the leaving-group in the unimolecular decomposition such as in (5). Metal-ion catalysis does not occur when the leaving-group is an aliphatic alcohol even when the oxocarbocation intermediate is quite table.^

The mechanism for the thermal disproportionation of quinone monoketals (6) involves reversible unimolecular dissociation into a radical pair followed by rate- limiting hydrogen transfer to give formaldehyde and 4-metho~yphenol.~

The sulphuric acid-catalysed hydrolysis of dihydropyranal acetals gives cy- clenones whereas the use of pyridinium p-toluensulphonate gives b-keto- aldehydes.'

As expected, the intermolecular addition of thiols to acyl nitriles shows much more negative volumes of reaction and activation compared with the intramolecular cyclization of a hydroxy-ketone to a hemiketal."

1 Reactions of Aldehydes and Ketones and their Derivatives 3

Substituent effects on the silver ion-promoted hydrolysis of thioacetals suggest that the rate-limiting step involves unimolecular cleavage of the metal-bound substrate (7)."

A set of rules to describe selectivity in acetal formation has been prepared.12 The reductive cleavage of homo-chiral acetals using Lewis acid-hydride systems

gives the opposite stereochemical outcome to that achieved using aluminium hydride.'

Oxocarbenium ions can be generated from acetals by electrolysis and trapped by silicon-mediated nu~leophiles. '~ The kinetics of homolytic fission of acetals in solution at high pressure have been re~iewed. '~ The mechanism of the liquid phase homolytic halogenation of acetals has also been reviewed. l h

Hydrolysis and Formation of Glucosides, Nucleosides, Oxazines, and Related Compounds

Non-enzymic Reactions

Under high pressure, the specificity of glycosylation is increased which is attributed to a shift in the equilibrium from the monocyclic glycosyl cation towards the bicyclic acyloxonium cation with increasing pressure."

Heptamolybdate ion is, thermodynamically, the most preferred species in aqueous solution among the numerous polyanions of molybdic acid. This ion catalyses the epimerization of u-glucose and racemization of D-glyceraldehyde. It is suggested that catalysis arises not simply from 0x0-complexation of the monosac- charide and molybdate ion but from stabilization of a 3-oxa-allylic cationic species.

The alkaline degradation of cytosine nucleosides involves predominantly deamination to the corresponding uracil derivatives. Substituent effects indicate that this occurs by direct displacement of the amino group by hydroxide ion attack at C(4). Uracil nucleosides are cleaved by complete fragmentation of the pyrimidine ring initiated by hydroxide ion attack on C(5) of the base." The reaction with 6- substituted purine nucleosides consists of three consecutive reactions: attack of hydroxide ion on C(8) followed by ring-opening of the imidazole to an imine- aldehyde (8) and anomerization of the aglycone;20,21 deformylation of (8); and finally cleavage of the N-glycosidic bond.

Carbon-bridged purine cyclonucleosides such as (9) show a slower rate of acid- catalysed hydrolysis, compared with that of 3-methyladenosine by a factor of over lo4. This increased stability is attributed to a combination of the positively charged nitrogen retarding protonation and strain in the oxocarbocation intermediate.22

The effect of substituents in the purine has been studied in the acid-catalysed hydrolysis of 2'-deoxyadenosines. Sites of protonation were determined using 15N NMR spec t ro~copy .~~

Ring-chain tautomerism in 1,3-oxazines (10) shows equilibrium constants which may be correlated with substituents in the aromatic residue by a Hammett pf-value of 0.76.24

4 Organic Reaction Mechanisms 1987

yy) HO k2 OH

(9)

Enzymic Reactions The incubation of 2-deoxy-2-fluoro-~-~-glucopyranoside (1 1) with P-glucosidase results in a time-dependent loss of enzyme activity. This is attributed to rapid loss of the leaving-group giving a relatively stable glucosyl enzyme intermediate.25

Solvent isotope effects and trapping experiments indicate that the mechanism of human liver a-L-fucosidase involves rate-limiting proton transfer and regeneration of the a-stereochemistry at the anomeric carbon. Initial-burst kinetics are consistent with either the formation of a glycosyl-enzyme intermediate or a tight ion-pair."

Isotopic labelling experiments indicate that sucrose synthetase does not catalyse the cleavage of the scissile carbon oxygen bond of the glucosyl substrate in the absence of fructose.27

Enzymatic hydrolysis has been used for the stereospecific synthesis of enan- tiomerically pure pyranosides by hydrolysis of epoxide precursors.28

The NAD +-glycohydrolase-catalysed hydrolysis of pyridine dinucleotides lac- king the normal amide residue at C(3) of the pyridine residue (12) proceeds with rate- limiting pyridinium-ribose bond fission. The maximal rates of the enzyme-catalysed reaction, k,,,, exhibit a large dependence on the leaving-group pyridine with a Br4nsted of - 0.9. The Br$nsted value for the pH-independent uncatalysed unimolecular reaction is - 1.1. These results are consistent with the previously proposed late transition state with considerable oxocarbocation character.29

Reactions and Formation of Nitrogen Derivatives, Schiff Bases, Hydrazones, Oximes, and Related Species

Both increasing pressure and solvent polarity decrease the rate of isomerization about carbon-nitrogen double bonds. This is not expected for a rotational


mechanism involving charge separation in the transition state. The mechanism of isomerization, even for strongly electron-withdrawing and -releasing substituents, appears to proceed by an inversion pathway (13).30

An inversion mechanism has been proposed for E/Z-isomerization of sulphony- limines on the basis of small substituent effects.31

1,3-Diazolidines (14) in trifluoroacetic acid are in equilibrium with the open-chain iminium ion (15) and, with no 2-substituent, 5-10% exists in the chain form. Aryl substitution in the 2-position increases this proportion to about 30% whereas N-alkyl substitution has little influence on the ring-chain equilibri~m.~'

Ph, ,NHPh CH

0

PhCONHN-CHAr I I

Ph-C-CMe,CO,Et

(17) (18) (19)

Imine-enamine tautomerism occurs in 1,2- (16) and 2,5-dihydropyrimidines and the position of equilibrium is strongly dependent on substituents at positions 4 and 6.33

Theoretical calculations predict that the nucleophilic addition of ammonia to formaldehyde is catalysed by one or two water molecules participating bifunction- ally in cyclic activated complexes. Catalysis arises from the increased strength of hydrogen-bonding between water and the zwitterionic tetrahedral intermediate.34

The terminal tertiary amino group in (dimethy1amino)alkylamines acts as an intramolecular general base catalyst in the transamination reaction of the diamines with N-benzylideneaniline (17).35

The aminolysis of the cr,cl-dimethyl-substituted keto-ester (18) proceeds by initial attack on the ketone carbon.36 The kinetics and effect of substituents on the reaction of amines with N-sulphonylbenzimidates in chlorobenzene have been rep0rted.j'

The rate constants for the first five steps of the condensation of glucose with valine have been determined as well as the identification of the many products formed from


the acid degradation of the Amadori ir~termediate.~' The mechanisms of heterocyc- lic ring-closures to give five- and six-membered rings have been reviewed.39

Visual pigment rhodopsins contain a Schiff base which is hydrolysed at some stage during the photochemical cycle after cis-trans photo-isomerization and then reformed with 1 1-cis-retinal. The intermediate carbinolamine formed during the hydrolysis of N-retinylidine-n-butylamine, a model for the rhodopsin chromo- phore, has been detected using stopped-flow techniques. Hydrolysis in neutral micelles exhibits general base catalysis and spectroscopic changes in micelles are related to those observed in the natural system.40

An extensive study of substituent effects on the hydrolysis of Schiff bases to substituted benzaldehydes and 4-aminobenzoic has resolved carbinolamine formation and breakdown into Hammett p-values of - 1.54 and - 0.63, respectively. Some interesting, even if speculative, intermediates have been p r~posed .~ '

The effect of substituents on the acid-catalysed hydrolysis of the imines (19) has been in~estigated.~' Carbinolamines are postulated intermediates in the base- catalysed hydrolysis of pyrrole-substituted i m i n e ~ . ~ ~ Thermodynamic activation parameters for Schiff base formation from pyridoxal phosphate and hexylamine have been reported.44 The kinetics and mechanism of the condensation of 2-hexosamines with pyridoxal phosphate has been investigated and changes in rate-limiting step with reaction conditions observed.45

The transamination reaction between pyridoxal and pyridoxamine Schiff bases has been shown to involve uncomplexed rather than metal-ion-complexed carbanions, as suggested p r e v i o u ~ l y . ~ ~

The reaction of chalcone with phenylhydrazine gives two products, differentiated by which nitrogen initially attacks the a,a-unsaturated ketone (20). The reaction at the NH, position, but not that at NH, is base-catalysed. Although not considered by the authors, it is conceivable that addition at NH is catalysed intramolecularly by the oxygen of the enolate anion acting as a general base (21).47

The formation of indoles from the reaction between arylhydrazones and phosphorus trichloride occurs by the formation of a diazaphospholine intermediate (22). It is proposed that acid-catalysed P-N fission occurs in (22) to give intermediates similar to those suggested for the Fischer indole rea~t ion .~ '

Intermediates formed in the reaction of hydrazine with B-keto-esters have been studied by I3C NMR spectroscopy. Nucleophilic attack occurs initially at the keto carbon followed by ring-closure of the hydrazones or enamine t a ~ t o m e r s . ~ ~

Pyrazoles are formed from the reaction of a-aroyl-a-bromoketene dithioacetals with hydrazine. Rearrangement occurs in the episulphonium ion formed from the initial hydrazine adduct by displacement of bromide by the thioether." Activation parameters and Hammett p-values have been reported for phenylhydrazone formation from substituted benzaldehyde~.~ '

Tosylhydrazones react with peroxy-sulphur compounds to give the corresponding carbonyl derivatives by, it is assumed, base-catalysed fragmentation of the oxaziridine intermediate (23).52

Sulphonyl isocyanates react with aldehyde hydrazones at either the carbon or nitrogen of the azomethine. The position of electrophilic attack is determined by the


nature of s u b s t i t u e n t ~ . ~ ~ The effect of substituents and solvent on the azo-hydrazone tautomeric equilibrium has been discussed.54

The intramolecular addition of an amino group to a carbonyl group generates the same carbinolamine tetrahedral intermediate (24) as that formed from the addition of hydroxide ion to a cyclic iminium ion. Unexpectedly, the cation (25) is in equilibrium with the protonated amino-ketone below pH 9.5. For electron- withdrawing substituents in the aromatic residue, the ring-opened form constitutes up to 14% of the mixture. Furthermore, in basic solution the predominant species is not the pseudo-base (24) but the ring-opened deprotonated amino-ketone (26).

During the biosynthesis of nicotinamide from tryptophan, the aldehyde (27) ring- closes to pyridine-2,3-dicarboxylate by a non-enzyme-catalysed pathway. At low buffer concentrations, the dehydration of the carbinolamine intermediate is rate- limiting whilst at high concentrations ring-closure becomes the slow step.56

R do

(24)

Ar 1 Me

(25)

H


Q-PNH \ A5 H

(33)

co; I

Amines add to the fused thiazolium salt (28) to form stable pseudo-bases (29) whereas the reaction with (30) leads to ring-opening giving the enamine (31). These, and other, annelation effects are rationalized by calculations of the charge distribution in the reactant thiazolium salts.57

The iminium ion (32) is the presumed intermediate in the ring-closure reaction of amino-indoles with aldehydes. A spiro intermediate (33) is then thought to be formed from enamine-type carbanion addition to the electrophilic carbon in (32). The stereochemistry of this addition step and the subsequent 1,2-shift controls the stereochemistry of the product.58 This attack at the indole 3-position in preference to the 2-position is yet another example of the supposedly unfavourable 5-endo-trig r ing -c l~su re .~~

The condensation of benzylidenemethylamine with homophthaiic anhydride gives substituted piperidinones. Substituent effects indicate that the reaction occurs by initial iminolysis (34) in which E- and Z-imines ring-close to cis- and trans- isoquinolines, respectively, by different rate-limiting steps.60

Deamination of 1-aminopurinium salts with methanolic ammonia occurs by a nucleophilic attack of ammonia on the cyclic iminium ion followed by a ring- opening and -closing process (35) as indicated by isotopic labelling experiments.61

The reaction of methylamine with 1,4-dialdehydes to give intermediate imines which ring-close to give iminium salts (36) may be related to the ‘hot taste’ of the dialdehydes.62

1 Reactions of Aldehydes and Ketones and their Deriautives 9

The reaction of 1,3-diones with chloromethylene-iminium salts probably proceeds by the intermediate formation of iminium cations.63 The kinetics of the cyclo- condensation of 0-phenylenediamine with benzil to give 2,3-diphenylquinoxaline have been reported.64

Iminium salts generated from primary amines and formaldehyde react with allylstannanes in protic media to give bishomoallylamines by a rapid aminomethano-de~tannylation.~~

The hydrolysis of the imine of the 1,Cbenzodiazepine (37) psychotropic drugs is thought to involve reversible ring-opening to the amino-ketone (38)66 and the kinetics of this process in acid solution to give the triazole benzophenone derivative have been reported.67

The reaction of hydroxylamine with 4-phenyl- 1,5-benzodiazepin 2-thiones and diazepinones (39) gives the 3-substituted amino-isoxazole (40). The proposed mechanism involves hydroxylamine attack at both unsaturated carbons followed by ring-closure.68

Analogous to the photo-hydration reactions of unsaturated carbon-carbon bonds, photo-hydrolysis of oximes occurs readily. Electronic excitation to the lowest singlet state (41) leads to a highly polarized state with an increase in the basicity of the oxime nitrogen. Oxazirane intermediates are formed from oximes in competition with the addition of water but not from oxime ethers, the reaction of which is not acid- or base-catalysed. Photolysis of o-hydroxy-substituted aromatic oximes gives substituted benzoxazoles in competition with acid-catalysed h y d r o l y ~ i s . ~ ~

The reaction of the carbinolamine (42) with hydroxylamine and other carbonyl reagents gives the expected ring-opened oxime and other products, respectively. This is contrary to an earlier report7'

Dialkyl acetals of N, N-dialkylformamide react irreversibly with amines to form amidines and although Hammett o-values have been reported there are few mechanistic data available for this reaction7'

Me I

" O R

Ar

(37)


(43) (44)

The nitrosation of ketones gives a-nitroso-ketones, usually existing as the oxime tautomers, and the acid-catalysed reaction proceeds by rate-limiting enol formation or nitrosation of the e n 0 1 . ~ ~

Substituted nitrosobenzenes react with formaldehyde in an acid-catalysed reaction giving N-phenylhydroxamic acids by initial nucleophilic addition to the aldehyde conjugate acid.73

There is no observed intramolecular general base catalysis by a pyridine residue in the metal-ion-catalysed hydrolysis of acetylpyridine ketoxime pyridine- c a r b ~ x y l a t e s . ~ ~

The reaction of chalcone with hydroxylamine in the presence of activated barium hydroxide as catalyst gives predominantly isoxazolines. It is suggested that the metal ion controls the cyclization process.75 The same catalyst is also effective in the Claisen-Schmidt condensation. The reaction appears to be an inter-facial solid- liquid process between the adsorbed carbanion and the ben~a ldehyde .~~

a-Bromoacetophenone oxime undergoes halogen substitution with cyanide to form an isoxazole by intramolecular attack of the oxime hydroxyl at the nitrile carbon.77 Benzamidoximes react with aldehydes to form 4,5-dihydro-1,2,4- oxadiazoles, presumably by ring-closure of the intermediate amidine.78

Borane reacts with chiral vicinal amino-alcohols to form oxazaborolidines (43) which do not reduce ketones by themselves. However, in the presence of a molar equivalent of borane-tetrahydrofuran, reduction is rapid and enantioselcctive even with a catalytic amount of (43). The selectivity is attributed to reduction occurring in the coordinated complex (44).79

f Reactions of Aldehydes atid Ketones and their Dcrivutiues I1

C-C Bond Formation and Fission; Aldol and Related Reactions

The reaction of a,a-dichloroacetoacetate with aldehydes in the presence of ethoxide ion gives epoxides. This novel reaction presumably involves a retro-acetoacetate condensation and subsequent dichlorocarbanion attack on the aldehyde. The tetrahedral intermediate anion (45) must then ring-close to the epoxide by displacing chloride.80

The general stcreochemical outcome of aldol reactions that E - and Z-enolates give preferentially anti and syn aldols, respectively, is usually rationalized by the chair-like transition state. Assuming that only directionality and frontier-orbital interactions are important in these reactions under kinetic control, it has been shown that the stereochemistry of the product may be deduced."

Geometrical equilibration of acyclic enolates from the often kinetically favoured E-isomer to the more stable Z-isomer (46) is a common method for controlling enolate stereochemistry. The effect of alkyl substituents on the equilibrium has been measured and compared with force-field calculations.''

The pyridine catalysed 2-E isomerization of arylideneflavanones occurs by initial addition of pyridine to the a,,&unsaturated ketone.83 The diastereoselectivity of the aldol condensation ofchiral glycol enolates is metal-tunable to give either syn or anti ad duct^.'^ The diastereofacial selectivity of enolates in aldol condensation can be varied by modification of a distant s ~ b s t i t u e n t . ~ ~

The enolate derived from the keto-pyranoside (47) undergoes aldol condensations in which the pyranoside core acts as a nucleophilic centre without b-elimination of the glycosidic methoxyl group and assists stereoselectivity.86

Sonication affects the stereochemical outcome of the Barbier reaction between 2- S-octyl halides, lithium and cyclohexanone. Ultrasonic waves are thought to affect single-electron-transfer at the solid-liquid i n t e r f a ~ e . ~ ~

The Michael reaction of ascorbic acid with acrolein yields the spiran (48) which on distillation loses water to give the previously described product. This is the ring- closed hemiacetal form of the expected hydroxy-aldehyde adduct."

0.- c1 I I

I RCH-C-COZEt

CI R'

OH I

OMe

HO

(47)


(49)

The first direct evidence of the intermediate Cram chelate of organometallics with a-alkoxy-ketones has been obtained using I3C NMR s p e c t r o s ~ o p y . ~ ~ Enantio- selectivity is observed in pinacol formation from camphor which is attributed to metal-ion co~rdination. '~

A model of the intermediate complex formed during the enantioselective addition of diethylzinc to benzaldehyde with chiral tridentate lithium complexes as catalysts is able to predict the observed direction of enantio~electivity.~'

Steric effects in the transition state have been used to account for anti-selectivity in the aldol reaction using organoboron corn pound^.^^ The hydrogenation of fl-keto- esters to P-hydroxy-esters in high enantiomeric purity can be achieved using substituted phosphines coordinated to ruthenium@) as a catalyst.93

The synthesis of secondary alcohols in enantiomeric excess can be achieved by converting aldehydes into asymmetric dioxolanones, imidazolidinones, and oxazolidinones followed by nucleophilic addition. Using chiral diols coordinated to titanium, enantioselective addition of nucleophiles occurs to aromatic aldehydes.94

Using chiral diamine ligands, asymmetric addition of arylmagnesium bromide to aldehydes gives carbinols in enantiomeric exces~. '~ The addition of E- and Z-crotylboronates to aldoximes shows unexpected diastereoselectivity with the E-isomer giving the anti and the 2-isomer giving the syn hydroxylamine p r ~ d u c t . ' ~ The addition of dialkylzincs to aldehydes catalysed by chiral pyrolidinyfmethanols gives products in up to 100% enantiomeric excess.97

As usual, there have been many other reports on the stereochemical outcome of carbon-carbon bond-forming reactions.'* The reaction of tetraacetylethene with dimethyltitanium dichloride gives (49) presumably via ring-closure of the intermediate carbocation (50).99

The equilibrium between the intermediates formed in the Reformatsky reaction of benzaldehyde and cyclohexanone with the bromo-zinc salt derived from y-bromo- senecionic acid has been investigated.'00

The relative rates of addition of methyltitanium reagents to cyclic ketones follow the unexpected order for ring sizes 5 , 6 and 7; aldehydes react about 500-fold faster than ketones.'" Aldol condensations initiated electrochemically are assisted by the presence of lanthanide(u1) ions.'02

Silyl enol ethers add to acetals, aldehydes, and a,P-unsaturated ketones in the presence of trimethylsilyl chloride and tin(I1) chloride. It is assumed that the tin@) polarizes the silicon, improving coordination to the carbonyl oxygen.lo3

The reaction between silyl-ketene acetals and benzaldehyde in the presence of titanium(1v) chloride is thought to proceed by the formation of an octahedral, six- coordinate metal complex.'04


The rapid addition of a tertiary amine to tin(1v) chloride-selenoacetal mixtures gives the relatively stable corresponding vinyl selenides, whereas the slow addition of the base gives aldol condensation products. The difference is attributed to the geometry of the vinyl-selenide formed.Io5

The rate of the intramolecular aldol reaction of (51) is second-order in S-proline but the enantiomeric excess increases non-linearly with the concentration of the base.Io6

Intramolecular cyclization of the C(3) carbon of dihydropyridine on an acetal carbon is catalysed by titanium tetrachloride and hydrogen chloride. The intermediate carbocation is presumably trapped by the enamine-type residue.lo7

Azoles react with ethylbromopyruvate oxime to give N-alkylated products of the azoles. Basic azoles, with pK, values above 4, react by an elimination-addition mechanism. lo

The condensation of salicylaldehyde and the piperidine dienamine of citral gives a stable hemiacetal (52) arising from initial &attack on the enamine."'

The regio- and stereo-selective em-alkylation of A2-isoxazolines is thought to be due to selective Z-aza-enolate formation followed by alkylation at the least hindered site. O

The high diastereoselectivity of the Grignard addition to triazole ketones (53) is rationalized in terms of azole coordination to magnesium.' '

0 cb 0

t Ph,PMe 0-

Et0,C D H

(55)

+ EtO,C--CH-PMe,Ph

There has been a further report that the reaction of Grignard reagents with benzil and benzophenone proceeds by an electron-transfer mechanism involving a radical


cation of the Grignard reagent.' l 2 Although this is disputed,Il3 the radical ion-pair dimer gives well-resolved ESR signals, the stability of which is related to the reduction potential of the ketone and the oxidation potential of the Grignard reagent. As well as the reduction product (benzoin), the reaction of benzil with Grignard reagents gives the C-alkylation (1,Zaddition) and 0-alkylation (1,4- addition) products. An electron-transfer mechanism is suggested involving a six- membered transition state (54). The C-alkylated product can result from recombin- ation of the radicals after their separation from the solvent cage.'I3

The addition of methyllithium and methyl-Grignard reagents to a,B-epoxy- aldehydes occurs stereoselectively to give products which correspond to 'non- chelation-controlled' addition to the aldehyde group.' l4

Dithioacetals give the corresponding coupled alkene by treatment with Grignard reagents in the presence of a nickel-phosphine complex. It is suggested that the Grignard reagent traps the intermediate formed from carbon-sulphur bond- cleavage.' '

Unsolvated dibutylmagnesium reacts with aldehydes in the presence of lithium alkoxide in methylcyclohexane to give not only the expected alcohol but also the ketone formed by oxidation.' l 6

Although phosphonium ylides react with carbonyl groups to form alkenes and a phosphine oxide (Wittig reaction), sulphonium ylides form the thermodynamically more stable oxirane and sulphide. The sulphonium ylides are more reactive and add more readily to carbonyl groups and theoretical calculations indicate that the major difference is the fate of the betaine and cyclic intermediates.'17

The stereochemistry and mechanism of the Wittig reaction have been reviewed.' ' * Linear-free-energy relationships for the Wittig reaction between substituted benzaldehydes and triphenylbenzylidenephosphines in methanol have been reported.' l 9 The negative volumes of activation for the Wittig reaction have guided the application of high pressure to the Wittig synthesis of tri- and tetra-substituted alkenes.' 2o

The stereoselectivity of the Wittig synthesis of P-unsaturated esters from carbohydrate-derived alkoxy-aldehydes depends on the substrate structure and the nature of the solvent. It is suggested that the alkoxy groups are involved in solvation of the phosphonium group.12'

Replacement of triphenylphosphine by P-diphenylphosphinopropranoic acid in the Wittig reaction increases the E-stereoselectivity with semi-stabilized ylides.'22

Magnesium and zinc oxides are efficient catalysts in the Wittig and Knoevenegel reactions because, it is suggested, the metal ion acts as an electrophile to the carbonyl oxygen and the oxide anion acts as a base catalyst in carbanion formation.' 2 3

a-Hydroxy-ketones undergo an accelerated Wittig reaction with stabilized phosphonium ylides to give E-trisubstituted alkenes; this is attributed to either intramolecular hydrogen-bonding to the carbonyl oxygen or to an interaction between the hydroxyl oxygen and the ylide p h o s p h ~ r u s . ' ~ ~

The betaine (55), generated from its precursor alcohol, decomposes stereospecifi- cally to the Z-alkene. The corresponding irreversible Wittig reaction of the ylide (56)

1 Reactions of Aldehydes and Ketones und their Derivatives 15

with cyclohexanecarboxaldehyde gives a 95 : 5 LIZ-alkene ratio. This kinetic preference for E-alkenes indicates that equilibrium arguments used to explain high E-selectivity for related Wittig reactions of stabilized ylides with aldehydes are not general.

The addition of sulphur ylides to aldehydes under heterogeneous conditions in the presence of small quantities of water gives epoxides, presumably by ring-closure of the intermediate (57).lZ6

Base-catalysed ring-opening of hydroxythietanes (58) to give carbonyl compounds occurs in competition with nucleophilic substitution. The sulphane reacts 4 x 104-fold faster than the corresponding acyclic derivative. These reactions are thought to be concerted (59) rather than generating the free ~ a r b a n i 0 n . l ~ ~

0- I

RCH-CHR

HO P

OH

Ar -C -CcI,

OH

I I Ar do

The alcoholysis of 2,2,2-trichloro- 1 -arylethanols is catalysed by amines which is thought to correspond to general acid-catalysed breakdown of the tetrahedral intermediate (60), although there must be doubts about the availability of electron density for the protonation step.'28

The mono- and di-anions of the ketone hydrate (61) both expel the trichloro- methyl anion.129

Nucleophilic addition to highly branched perfluoro-ketones invariably leads to carbonyl carbon-carbon bond cleavage, the leaving-group carbanion being stabilized by the electron-withdrawing fluorines. Another driving force may be steric hindrance in the tetrahedral intermediate since there is no evidence of hydrate formation of the fluorinated ketone in water.'30


Substituent effects in the decomposition of benzoins in sodium methoxide- methanol indicate that the rate-limiting step in this carbon-carbon bond-cleavage reaction is the addition of methoxide ion to the cdrbonyl group.13'

A suitably placed carboxyl group facilitates carbon-carbon bond-cleavage in a cyclopropane activated by a carbonyl group.13' The C-C bond cleavage of non- enolizable a-phenyl ketones with sodamide replaces the benzoyl residue with hydrogen with retention of configuration. 1 3 3

The ring-cleavage of the furandione (62) by amines is general base-catalysed in dioxane. Solvation effects on the rate of proton transfer have been considered in determining the rate-limiting step. ' 34

The one-carbon ring-expansion reactions of bridged bicyclic ketones have been reviewed.'

Other Addition Reactions

The rate of oxygen exchange in a-methyl-substituted ketones (63) lends support to the suggestion that the interaction between the electron pair on an incoming nucleophile and the antibonding orbital of the antiperiplanar C- bond is greater for C-C compared with C-H bonds.'36

Orbital arguments and torsional effects'37 suggest that nucleophilic attack on cyclohexanone occurs preferentially from the axial direction but this is frequently overwhelmed by steric effects. Support for this claim comes from the addition of small nucleophiles such as the acetonitrile carbanion which shows a preference for axial attack.'38

Theoretical calculations on the trajectory of nucleophiles attacking chiral aldehydes and ketones (64) support the idea that diastereofacial selectivity arises from steric effects; when R is methyl or ethyl the trajectory is more on the R side of the normal plane than on the phenylethyl side; when R is t-butyl the approach trajectory of the attacking nucleophile is on the phenylethyl side of the normal plane.' 39

The suggestion that nucleophiles attack carbonyl groups along a trajectory that is displaced from the normal plane in the direction of the sterically less-demanding substituents attached to the ketone appears to be applicable to additions to chiral thionium ions.14*

It has been suggested that the a-effect, the enhanced nucleophilicity of a base possessing an adjacent heteroatom with an unshared electron pair, arises from a stereoelectronic effect. In the addition of phosphite nucleophiles to carbonyl groups there can be more oxygen lone pairs antiperiplanar to the newly formed phosphorus-electrophile bond in acyclic compared with cyclic ph~sphites. '~'

The theoretical treatment of the stereoselective nucleophilic addition to carbonyl compounds has been reviewed,'42 as have steric effects'43 and product stability as a factor in kinetically-controlled addition reaction^.'^^

The hydration equilibrium of aliphatic aldehydes is reduced when the aldehyde is bound to cationic and anionic micelles because of a decrease in the water activity at


the micellar surface. The acid-catalysed hydration rate is increased by the presence of anionic micelles but decreased by cationic ones. ' 45

The hydration and acid equilibria for 5-deoxypyridoxal have been reported.'46 Dehydration of formaldehyde hydrate is important in the kinetics of formation of the sulphonic acid a d d ~ c t . ' ~ ~

The kinetic hydrogen isotope effect for the intramolecular hydride transfer in the glyoxal-glycollic acid rearrangement is calculated to be larger than that for analogous intermolecular hydride transfer. This is contrary to the general expect- ation of non-linear hydrogen transfer and, it is suggested, proton and hydride transfer may differ r a d i ~ a 1 l y . l ~ ~

Intramolecular hydride transfer in the hydroxy ketone (65) shows a primary kinetic isotope effect k,/k, of 3.4 for the migrating hydrogen which agrees well with the calculated value.'49

The borohydride reduction of the propellane carbonyl group in an aqueous solution of a cationic surfactant occurs not only with the expected rate enhancement but with a stereochemical preference which is attributed to the mode of binding the substrate to the micelle.' 5 0

The stereoselective reduction of enol acetates to alcohols by lithium aluminium hydride is attributed to metal-ion coordination to the ketone intermediate.' '

0

Me hie

(63)

R Me

M e 0


The enzyme mandelonitrile lyase catalyses the enantioselective formation of cyanohydrins from aldehydes in organic solvents because the rate of the normal chemical reaction is suppressed.' 52

Cyanide ion catalyses the dioxygen oxidation of a-keto-aldehydes by intermediate formation of the cyanohydrin anion which transfers an electron to oxygen.'53

Nucleophilic addition of cyanide anion to 2-benzopyrylium salts (66) gives stable adducts.' 54

A limited kinetic study of the conversion of pyrylium cations into pyridinium derivatives with primary amines has led to speculation about the rate-limiting step, with little supporting evidence.lS5

The reaction between nitroalkanes and o-quinones in carbonate buffers gives cyclic acetals (67). The nitroalkane anion is thought to add to the carbonyl oxygen of the o-quinone to give (68) which subsequently ring-closes to the product. There are several other pathways by which the intermediate (68) could be generated.'56

Enolization and Related Reactions

There have been more reports of stable enols derived from sterically-hindered ketones.' 57 The photochemical generation of vinyl alcohol in aqueous solution has allowed the determination of the pK, (10.5) for the enol and the pK, (16.7) of acetaldehyde. Kinetic solvent isotope effects on both enolization and ketonization suggest that the reaction occurs by a stepwise rather than a concerted mechanism. Combined with other data for base-catalysed enolization, there appears to be an imbalance in the transition state with charge delocalization in the enolate anion as indicated by Br@nsted a value, lagging behind proton transfer to the base, as indicated by the Br@nsted D value.lS8

There is a remarkable linear correlation between the pKeno, values for sterically hindered ketones in hexane and those for substituted ketones (69) in water. Although the data cover a limited range of equilibrium constants, the linearity of the correlation must reflect a proportionality between substituent effects in both systems.' 59

The enol and hydrate are the dominant forms of 9-formylfluorene (70) in aqueous solution and the unstable aldehyde is generated from the reaction of its ethanethiol hemiacetal with iodine. The stability of the enol is indicated by the pK, of 6.2 for the aldehyde (70). ' 6o

The equilibrium constants for the equilibrium between enols/alcohols and enol

Documents

ORGANIC REACTION MECHANISMS 1987...Ring-chain tautomerism in 1,3-oxazines (10) shows equilibrium constants which may be correlated with substituents in the aromatic residue by a Hammett