The Synthesis and Fine Structure Studies of Polycyclic Aromatic Hydrocarbons

The main part of this thesis has dealt with benzylic coupling in a large number of methyl derivatives of polycyclic aromatic hydrocarbons. It has been shown previously that a coupling constant of ~1.0 Hz. between the methyl protons and the ortho aromatic proton is indicative of a high degree of localisation and double bond character in the intervening bond. This idea has been extended by examining a series of symmetrical 6-dimethyl derivatives which involved synthesising 2' ,3'-dimethyl-1,2;6,7-dibenzo-pyrene, (I) and 6,7-dimethy1-1,2:3-4-dibenzanthracene (II). 9,10-Dimethyl-3,4-benzotetraphene (III) was obtained as a byproduct in the latter synthesis. The couplings in these derivatives add further evidence to the idea that there is only one true double bond in each ring which occupies the lowest energy orbital. The remaining 4 pi-electrons are divided into two higher energy levels with one pair remaining delocalised within the ring,and the third pair, in the highest level, delocalised outside the ring and capable of migrating from ring to ring. An annellation effect is thus observed in passing from o-xylene to 2,3-dimethylanthracene (IV)(V)(VI) which can be explained by the increased linear annellation reducing the influence of those delocalised pi-electrons which are free to migrate, as they are spread over more rings. The distribution of the true double bond can also be considerably affected by substitution in the ring. This is reflected by considerable -changes in the magnitude of the benzylic coupling and supports the additivity of only one true double bond in each ring. The electronic arrangement in pyridazine is in the opposite sense to naphthalene and anthracene. There is a high degree of double bond character in the 4,5 position and a correspondingly low value in the 3,4 and 5,6 positions. This is borne out on examination of a number of methylpyridazines and adds further evidence to the above ideas. Finally, the unusual asymmetry of the methyl resonance in alpha-methyl derivatives was investigated. During this work an explanation was published on the basis of computer analysis of the methyl signal and it appears that the asymmetry is inherent with this combination of couplings to the ortho, para and meta protons on the ring. The apparent lack of long-range coupling in aromatic hydrocarbons has been attributed to 2nd order interactions. In this work halogen derivatives of anthracene and tetracene were examined. Decoupling indicates the absence of para coupling between the protons on non-terminal rings, although there is substantial trans-peri coupling between the protons of a terminal and non-terminal ring. However a methyl group on a non-terminal ring is found to he strongly coupled only to the para proton with J= 0.8-1.0Hz. This value is as large as the coupling constant observed for the localised double bond in 9-methylphenanthrene 1.0 Hz. This evidence is compared with other methyl derivatives and several possible explanations are discussed. The reductive dimerisation of 6-naphthanthrone yielded the known hydrocarbons, dinaphtho-(7', 1' : 1,13): (1", 7" :6, 8) -peropyrene , dinaphtho- -peropyrene and 2,3-trimethylenepyrene along with two new hydrocarbons (VIl) and(VIIl). Circobiphenyl (VII) is a highly condensed hydrocarbon which can be considered as a double coronene. The U.V. spectrum is similar to coronene, however its low solubility has proved a stumbling block in efforts to obtain the n. m. r. spectrum. It was hoped that the protons of this molecule would resonate at low field thus indicating a "superaromaticity" effect analogous to coronene. A new purple hydrocarbon was also obtained and U.V. evidence favours the structure dibenzisoviolanthrene (VIIl). Application of Robinsons aromatic sextet to this structure does not indicate any localised dienophilic regions and therefore this molecule should not react with maleic anhydride. Contrary to expectations, two-fold addition of maleic anhydride does occur in a similar manner to 2,3:4,5:8,9:10,11-tetrabenzoperylene and on this basis the possibility of structure (IX) cannot be excluded. However closer examination of structure (IX) shows that a complete Kekule structure is not possible, despite the fact that alternant marking predicts an equal number of marked and unmarked carbon atoms. The former finding leads to the theoretical prediction of a diradical or higher structure with an inherent instability as shown by triangulene, yet from the alternant marking sbbeme (IXa) a bond across the middle ring involving electrons of opposite spin might be possible. In order to test the above predictions and to try and clarify the reactivity of the purple hydrocarbon with maleic anhydride, an unambiguous synthesis of (IX) was attempted. A series of examples which show the asymmetric annellation of two diphenyl complexes to an acene have been built up over the years. Two of the remaining members of this series are 1,2:3,4:11,12: 13,14-hetrabenzheptacene (x) and 1,2:3,4-dibenzhexacene (xi). A synthesis of the former compound was undertaken to increase the scope of the above series and in an effort to obtain a reasonably stable heptacene derivative