A variety of methods has been employed in the investigation of the number of rotational isomers of a molecule, their relative energies, and the heights of the potential-energy curves for internal rotation. These include thermodynamics (heat capacities), molecular orbital calculations, electron diffraction, microwave and vibrational spectroscopies, relaxation measurements, and NMR spectroscopic studies. The method of choice' for any particular molecule depends upon the molecular structure, the magnitude of the barrier to internal rotation, and the approximations of the model to be applied.For barriers of <3 kcal/mol the most accurate determination of barriers to methyl group rotation is probably afforded by high-resolution microwave spectroscopy (gas phase). Even this approach contains approximations, such as semirigid Hamiltonians and ignored vibrations/torsions, which may lead to unknown errors in the final value for the barrier.2a The method is applied to fairly small molecules, and a molecule as large as phenylethane still presents the microwave spectroscopists with a formidable challenge. and phenol3 have been studied.For rotational barriers of larger magnitude the method of choice is probably dynamic NMR spectroscopy (DNMR). It is applicable to barriers as large as 30 kcal/mol and as small as 8-10 kcal/mol. The lowest barrier found in this way4 is 4.2 kcal/mol. DNMR makes use of the fact that different conformers may display different chemical shifts and/or coupling constants. The conversion rate between the conformers is studied as a function of temperature. Because DNMR depends upon the dephasing of magnetization as nuclear spins are transferred between sites of different Larmor frequencies ( T2 effects), this technique becomes more difficult to apply as the magnitude of the barrier decreases. The dephasing process becomes too fast to measure accurately.As NMR can be applied t o relatively large molecules in solution, it is desirable to extend it to the measurement of barriers of less than 4 kcal/mol. Nuclear, particularly 13C, spin-lattice relaxation rates are widely Bill Parr was born in England in 1948 and studied for his BSc. at Brighton Polytechnic before going to the University of York to work with R. 0. C. Norman for his doctorate on the oxidation of organic compounds by gold(II1) and palladium(I1) salts. In 1974 h e moved to the University of Manitoba, where he spent 2 years at postdoctoral work. After returning to England and a year at the University of Warwick as a postdoctoral fellow working on stopped-flow FT NMR, he joined MRPRA a s project leader in charge of oxidation studies.Ted Schaefer was born in Gnadenthai, Manitoba, in

We did not find i t so.the two metal oxidants but the yields were significantly higher with gold(II1).

The products of oxidation of a variety of olefins by tetrachloroauric acid in methanol are reported. They are satisfactorily accounted for by a scheme in which gold(ll1) acts as an electrophile to give an organometallic adduct: this breaks down by heterolysis of the C-Au bond, accompanied by competition between rearrangement of a neighbouring substituent and uptake of a nudeophile. In all cases a ctosesimifarity w a s observed with the products obtained from oxidations by thallium(lll) and lead(1v) : where divergences occur they can be accounted for in terms o i either the different ligands about the metal atom or the different leaving-group abilities of the metal substituents.IN recent papers 1-3 we compared the organic products formed in the oxidation of typical alkenes by mercury(II), thallium(m), and lead(rv) salts, the primary emphasis being on the mechanistic aspects of the reaction. In most cases a mixture of products is obtained but, under certain conditions, the reaction can be made synthetically useful.3-5 Usually the initial step is electrophilic attack on the double bond by the oxidant, which is followed by heterolysis of the carbon-metal bond in the resultant labile organometallic species. We sought other metal oxidising agents with comparable properties, in the hope that they might prove mechanistically and synthetically interesting. The metal must be a twoelectron oxidant, so that electrophilic metallation will occur. Unfortunately such elements are rare. Apart from the above three, the only other obvious candidates are platinum, palladium, gold, and tin. The first two, especially palladium( 11) , have been extensively investigated, but our ability to predict and control the products of their reactions with most alkenes is still limited; certainly the characteristics normally associated with electrophilic addition are not particularly evident with these two metals. In this paper we consider the usefulness of gold as a two-electron oxidant of alkenes.The first olefin-gold(II1) complex, that with cycloocta-l,.fi-diene, was identified as recently as 1964.' Such complexes are even less stable than those of gold(I), which themselves are difficult enough to isolate, decomposing to give the free alkene at temperatures little above ambient.8 Indeed there appear to be no authenticated examples of complexes of gold(II1) with simple alkenes, though mixed gold(I)-gold(m) complexes have been ident ified.9The few studies to date of the reactions of gold oxidants with alkenes have been primarily directed at organometallic aspects of the process, but two reports have appeared of the organic products. The first was a patent covering the production of aldehydes and ketones by reaction with tetrachloroauric acid.1° The second

The proton magnetic resonance spectra of some benzylfluoride derivatives yield long-range spin–spin coupling constants between ring protons and protons or fluorine nuclei in the fluoromethyl group. In conjunction with the eigenfunctions for a hindered twofold rotor, the couplings over six bonds are used to show that in 3,5-dichlorobenzylfluoride in solution the C—F bond prefers the benzene plane by 260 ± 50 cal/mol; in close agreement with ab initia and MINDO/3 molecular orbital calculations. The latter method suggests that in a conformation in which the C—F bond lies in a plane perpendicular to the benzene ring, the C – C – F angle reduces to 107.2° and the C – C – H angles become 116.1°, perhaps due to increased conjugation of the C—F bond or fluorine atom with the π electrons of the ring. The observed barrier is presumably a delicate balance between steric interactions, hyperconjugation or p–p conjugation effects, and dipole–dipole interactions between polarized bonds.

No abstract