Aromatic and Heteroaromatic Chemistry Volume 3

A Review of the Literature Abstracted between July 1973 and June 1974

By C. W. Bird, G. W. H. Cheeseman

The Royal Society of Chemistry

Copyright © 1975 The Chemical Society
All rights reserved.
ISBN: 978-0-85186-773-1

Contents

Chapter 1 Ring Systems of Topical Interest By P. J. Garratt, 1,
Chapter 2 lntrarnolecular Cyclizations By A. W. Somerville, 73,
Chapter 3 Condensation Reactions By P.A. Lowe, 112,
Chapter 4 Cycloaddition Reactions By G. V. Boyd, 145,
Chapter 5 Ring lnterconversions By A. J. Boulton, 183,
Chapter 6 Electrophilic Substitution on Carbon By R. Taylor, 220,
Chapter 7 Electrophilic Substitution on Heteroatoms By E. F. V. Scriven, 262,
Chapter 8 Nucleophilic Substitution By G. B. Barlin, 275,
Chapter 9 Aromatic Substitution by Free Radicals, Carbenes, and Nitrenes By S. R. Challand, 301,
Chapter 10 Addition Reactions By G. V. Boyd, 319,
Chapter 11 Ring-cleavage Reactions By T. L. Gilchrist, 363,
Chapter 12 Reactions of Substituents By P. D. Magnus, 377,
Chapter 13 Porphyrins and Related Compounds By K. M. Smith, 409,
Chapter 14 Naturally Occurring Oxygen-ring Compounds By R. D. H. Murray, 431,
Chapter 15 Other Naturally Occurring Compounds By J. R. Lewis, 454,

CHAPTER 1

Ring Systems of Topical Interest

BY P. J. GARRATT

1 Introduction

On taking over the writing of this chapter from the Senior Reporters, the present author has adopted the general organization which they have used for the previous two volumes. Examination of Chemical Abstracts Volumes 79 and 80 indicates that, despite the strictures of one reviewer of Volume 1 of this series, the activity of chemists in this area continues at a high level. Binsch, in an elegantly written article entitled ‘Aromaticity — An Exercise in Chemical Futility?’, cogently argues in favour of retaining this amorphous concept, and against those who prefer to remove anomalies in the interest of classification and general rules. As he points out, ‘the very essence of chemistry consists in complexity’, a statement that the proponents of ‘modern’ school chemistry might well consider. The problem of aromaticity has also been discussed by Hauptmann. Hobey has presented an FEMO model supporting the Kruszewski-Krygowski index of aromaticity. Theoretical justifications for Fries’ rule and for Dewar’s resonance-energy model have been advanced. Fries’ rule predicts that those canonical forms containing the most benzene rings will be preferred, a sentiment with which most experimental chemists intuitively agree. This rule is extended in that the larger the number of (4n + 2)-and the smaller the number of 4n-rings in a canonical form the more it will be favoured. The use of diamagnetic susceptibility anisotropy (DSA) and magnetic susceptibility anisotropy (MSA) as criteria for aromaticity has been examined, and it is suggested that both are probably invalid. Benzene, particularly at the undergraduate level, is usually treated as a rigid hexagon, at least as far as its chemical properties are concerned. Attention has been drawn to the ease with which benzene undergoes out-of-plane distortions. The experimental findings and a theoretical treatment are given, and it is suggested that mono-and poly-nuclear aromatic hydrocarbons are best treated as flexible systems capable of deviations of 5 — 20° from the plane in the ground state.

The Proceedings of the International Symposium on Aromaticity held in Jerusalem have, somewhat belatedly, been abstracted. Breslow has described the efforts of his group towards clarification of the problem of antiaromaticity. Seven-membered conjugated and heterocyclic rings, substituted cyclopropyl cations, cyclopropenones, the reactions of the radical anions and dianions of aromatic hydrocarbons, and arynes have all been comprehensively reviewed. Volume 4 of ‘Carbonium lons’ contains relevant articles on aryl, tropylium, bridgehead, and degenerate members of this type. Vogel has reviewed bridged Hückel aromatic systems.

2 Valence Isomers

Theoretical studies have been carried out on Dewar-benzene (1), benzvalene (2), and prismane (3). Bond lengths and bond angles were calculated and found to be in good agreement with the known values. The dipole moment of benzvalene was derived, and it is in good accord with the value obtained from the microwave spectrum. Dewar-benzene is predicted to have a very small (<0.04 D)dipole moment. The i.r. and Raman spectra of the hexamethyl derivatives of (1) and (3) have been analysed. A full account of the microwave spectrum of benzvalene has appeared, including the results of isotopic labelling, and the structure derived for (2) is shown in the Figure.

The ready availability of (2) by the Katz synthesis has stimulated interest in the chemistry of this system. Katz and Nicolaou, in an investigation directed towards an alternative synthesis of prismane, have studied the reactions of (2) with chlorosulphonyl isocyanate and sulphenyl halides (Scheme 1). Chlorosulphonyl isocyanate gave (4) and (5) (3:1), which can be converted into the corresponding amides (6) and (7) by hydrolysis. The reaction involves dipolar attack on the double bond, followed by Wagner–Meerwein rearrangements and ring closure. Reaction of benzenesulphenyl chloride again occurs at the double bond, but Wagner–Meerwein rearrangement does not occur and (8) is obtained, which on oxidation and dehydrochlorination gives phenyl staurophenyl sulphone (9). The distaurophenyl sulphone (11) was prepared by a similar route from (10), which was obtained by treating (2) with sulphur dichloride. Cristl has also examined the reactions of (2) with a number of 1,3-dipolar reagents, and finds that these add to the double bond. For example, trimethylbenzonitrile oxide gave (12), and diazomethane gave (13). Dibromocarbene added to give (14).

Irradiation of 1,2,4,5-tetra(trimethylsilyl)benzene (15) gave the benzvalenes (16) and (17), together with the 1,2,3,5-benzene isomer (18) and the fulvene (19). The benz-valenes appear to be primary photoproducts.

The rearrangement of Dewar-benzene and its derivatives into benzene requires a high activation energy. It has now been shown that the activation energy is sufficiently large that some (one in 103 — 104) of the benzene molecules are produced in the triplet state. Addition of substituted anthracenes, particularly 9,10-dibromoanthracene, to the reaction leads to weak emissions characteristic of the anthracene fluorescence. The rate of rearrangement of the pentafluoro-Dewar-benzene (20) into the corresponding benzene derivative (21) is increased when R is an electron-withdrawing group.

Hexamethyl-Dewar-benzene (22), on bromination followed by treatment with lithium aluminium hydride at -70°C, gave the dihydroprismane (24), probably via the cation (23). A full account of the preparation of the Dewar-pyridine (25) and azaprisinane (26) derivatives has appeared.

The valence isomer (27) of benzo[b]thiepin has been prepared, which rearranges to benzo[b]thiepin (28) on treatment with a rhodium(I) catalyst. Compound (27) rearranges to the tricyclic isomer (29) thermally, on photoirradiation, or in the presence of silver(I) ions.

Srinivasan has shown that the (CH)10 hydrocarbon (30) disproportionates thermally to naphthalene and the dihydro-derivative (31). The loss of hydrogen (or D) from (30) was shown to be nearly random, and an intermediate with all carbons equivalent is required, possibly all-cis-[10]annulene (32).

3 Polybenzenoid Compounds

Pyrene epoxide (34) has been prepared by treatment of phenanthrene-4,5-dicarboxaldehyde (33) with tris(dimethylamino)phosphine. On photoirradiation (34) gave the oxepin (35). Pyrolysis of the tosylhydrazone sodium salt (36) at 410°C gave a low yield of the hydrocarbon (37). Compound (37) could also be obtained by pyrolysis of (38), and the carbene (39) is probably involved (see Vol. 2, p. 56).

A new synthesis of decastarphene(3,3,3) (42) has been reported involving reduction and dehydration of (41), which was prepared by trimerization of the anthraquinone (40).

Circobiphenyl (see Vol. 2, p. 4) has been shown to have an E-type delayed fluorescence.

4 Helicenes

The heterohelicene (43) has been prepared by a conventional synthesis. On treatment of (43) with sodium tetrachloroaluminate the dehydrohelicene (44) was obtained. The heterohelicene (46) has been prepared by treatment of the dialdehyde (45) with benzylamine in the presence of sodium dithionite. A number of heterohelicenes have been prepared by the photocyclization route. Irradiation of (47) thus gave (48). It is reported that (48) and related compounds readily lose hydrogen to form the corresponding dehydrohelicenes.

A full account of the preparation of (49) has appeared. The two diastereomeric forms were isolated, the lower-melting diastereomer being transformed into the higher-melting on heating. It is suggested that the lower-melting form is the racemic compound, and some experimental evidence is given to support this view. Hexaheliceno[3,4-c]-hexahelicene (51) has been prepared in 15% yield by photoirradiation of (50). Only one diastereomer was obtained, and it is suggested that this is the racemic form.

5 Sterically Overcrowded Molecules

The n.m.r. spectra of the phenanthrenes (52a — d) (see Vol. 2, p. 6) have been used to test the classical and quantum-mechanical ring-current models. The X-ray structural data on these compounds are available, and the relative positions of the protons and the rings are thus known. For above-ring protons the classical Johnson-Bovey treatment is superior to the quantum-mechanical Haigh-Mallion–McWeeney method, the reverse of the finding for in-plane protons.

The biphenyl derivative (53) has been resolved. Heating the (+)-enantiomer at the melting point or to 65°C in solution converts it into the racemic mixture owing to equilibrium between (53a) and (53b).

6 Bridged Aromatic Compounds

The photoelectron spectra (p.e.) of a number of cyclophanes have been observed. Boschi and Schmidt have reported that strong transannular interactions occur in compounds (54) — (58), the interaction for (56) and (57) being more than ten times as large as that estimated from the u.v. spectra. Whereas in (54) — (56) the interaction occurs over all of the carbon atoms of the benzene rings, in (57) and (58) it is restricted to C-8 and C-16. Detailed accounts of the p.e. spectra of (54) and (55) have been reported. Comparison of the p.e. spectra of (56), (59), and (60) shows that substitution by fluorine ‘turns off’ the through-bond interaction by lowering the energy of the σ-orbital out of reach of the interacting π-orbital.

Gundermann and Röker have observed chemiluminescence in the cyclophanes (61) and (62). Reaction of (61) with oxygen in the presence of DMSO and potassium t-butoxide gave a weak fluorescence at 395 nm, and under the same conditions (61) gave a stronger fluorescence at 450 nm. The chemiluminescence appears to be a function of the transannular interaction in the two systems, since the hydrazide (63) gave only an extremely weak fluorescence.

[2,2]Paracyclophanes. — Solvolytic ring expansion of [2,2]paracyclophane (64) to the [2,3]paracyclophane (65) has been reported.48 Rebafka and Staab have prepared the interesting quinone (67) and quinone-hydroquinone (68) by the route shown in Scheme 2. Compound (68) gives a wine-red solution with the long-wavelength maxima at 515 nm (ε 170), but the intramolecular redox reaction does not appear to occur. The compound is assigned the trans stereochemistry shown, which may account for the lack of oxidation–reduction. The precursor (66) was also obtained in only one isomeric form.

A superior synthesis of [2,2] (2,5)-pyridinoparacyclophane (70) involves the photo-irradiation of (69) in the presence of a trialkyl phosphite. Compound (70) could not be obtained from the disulphone corresponding to (69).

[2,2]Metaparacyclophanes. — The barrier to ring-flipping in [2,2]metaparacyclophanes [(71)[??](72)] has been further studied (see Vol. 2, p. 10) by examination of the temperature dependence of the n.m.r. spectra of compounds with substituents at position 5.

The rate of ring-flipping was found for different substituents Y. A Hammett plot of σP against log ky/kH was not linear, but gave a smooth curve. Both electron-releasing and electron-withdrawing groups were rate-retarding and although the 5-position is remote from the site of flipping, the effects were quite large (NH2, the largest, showed a five-fold retardation in rate). A full account of the synthesis of [2,2]metacyclophane-1,9-diene (73a) has now appeared, together with the preparation of a number of derivatives (73b — d). The aldehyde (73d) has an absorption in the electronic spectrum at 349 nm (ε 786) and the aldehydic proton appears at τ 1.02 in the n.m.r. spectrum. Both the long-wavelength absorption and the high-field value of the aldehydic proton resonance can be attributed to an interaction between the aldehyde group and the benzene ring to which it is not covalently bonded. In constrast to [2,2]metaparacyclophane, the diene (73a) does not rearrange to the corresponding [2,2]metacyclophane-1,7-diene on photoirradiation.

[2,2]-2,6-pyridinoparacyclophane (74) has been prepared from the corresponding disulphide (seep. 12) by photoirradiation, and the diene (75) has been obtained by the Stevens–Hofmann sequence from the same precursor. Hydrogenation of (75) also gave (74), but a third method, pyrolysis of the disulphone, is apparently the most convenient. The n.m.r. spectrum of (74) at low temperature shows two types of para-bridged ring proton, but the barrier to ring-flipping is low (10.7 kcal mol-1). In contrast, the n.m.r. spectrum of (75) shows only one type of para-bridged ring proton down to -110°C, suggesting either a very small barrier to ring-flipping or that the pyridine ring is perpendicular to the benzene ring. An X-ray crystallographic analysis of (75) shows that, at least in the crystalline state, the molecule has the perpendicular conformation. The fluoroborate salts of (74) and (75) have been prepared, and the n.m.r. spectrum of the latter is again temperature-independent down to -100°C. A symmetric structure for the salt is again a possibility, indicating an interaction between the anion and the nitrogen positive pole via the benzene ring.

[2,2]Metacyclophanes. — The [2,2]metacyclophanes (76a and b) have been prepared by photoirradiation of the corresponding disulphide in the presence of trimethyl phosphite. [2,2](3,5)pyridinometacyclophane (77) has been prepared by the same procedure using triethyl phosphite.

Other Bridged-ring Systems. — The record for the smallest paracyclophane has now gone to the Princeton group, who obtained [6]paracyclophane (79) by pyrolysis of the lithium salt (78), a method they had previously used to prepare [7]paracyclophane (see Vol. 2, p. 11). The electronic spectrum of (79) shows absorptions at 212 nm (log ε 43), 253 nm (4.0), and 296 nm (2.8), and in the n.m.r. spectrum the benzene-ring protons appear at τ 2.83, H-1at 7.51, H-2 at 8.85, and H-3 at 9.67. The benzene ring is presumably considerably deformed out of the plane, but the position of the ring protons and the shielding effect on the methylene protons clearly indicate that it still supports a ring current.

Pyrolysis of dicyclopropylcyclohexa-1,4-diene (80) in the presence of a butadiene gave the [8]paracyclophanes (81) and (82). In the n.m.r. spectrum the olefinic protons of the trans-double-bond isomer resonate at higher field (τ 5.7 — 6.3) than those of the cis-isomers (τ 5.3 — 5.7). The reactions appear to proceed via a biradical intermediate, a CIDNP effect having been observed.

The reactions of the syn-and anti-substituted [10](2,4)pyridinophanes (83) and (84) with a variety of reagents have been compared. The substituents on the methylene chain are extremely inert, but when reaction does occur, for example in the hydrolysis of the tosylates (83a) with 90% formic acid-water to give a mixture of the alcohol (83b) and formate (83c), the configuration is completely retained.

The inversion of the [7](2,6)pyridinophane (85) has been studied by standard n.m.r. techniques. At low temperature two signals for the methyl groups are observed at τ 8.96 and 9.44, and the value of ΔG≠ is low (11.8 kcal mol-1).

The paracyclophanes (89) were prepared from the sulphides (86) by the Stevens rearrangement to (87), which, in the preferred method, were converted into the corresponding N-toluene-p-sulphonylsulphilimines (88) and pyrolysed in boiling xylene to give (89). The compounds with n = 3 — 6 have electronic spectra of a benzenoid type, whereas the systems with n = 9 or 10 have cis-stilbenoid-type spectra. The compounds with shorter methylene bridges also showed a greater preponderance of the cis-dibromide on bromination.

A comparison of the rotation of the benzene rings in the paracyclophanes (90) and (91) shows that it is slower in the compounds with methoxy-substituents (90) than in those with methyl substituents (91). In the cases chosen, the total bulk of the substituent is important, and thus OMe is larger than Me, opposite to the more usual finding for the steric effect of these two groups.

(Continues…)Excerpted from Aromatic and Heteroaromatic Chemistry Volume 3 by C. W. Bird, G. W. H. Cheeseman. Copyright © 1975 The Chemical Society. Excerpted by permission of The Royal Society of Chemistry.
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.