The Anderson—Weiss formalism is used to develop a general theory for calculating the effects of chemical exchange and spin coupling on echo amplitudes in Carr—Purcell pulse sequences on spin systems in the liquid phase. The theory is shown to involve the matrix integral solution of the Hahn—Maxwell—McConnell equations, generalized to include spin coupling, and of Alexander's equations for the relevant density matrix elements, subject to the boundary conditions imposed by the pulses. Some specific systems are treated in detail, and closed formulas are given for the decay of the echo train in several situations, including the coupled and uncoupled AB systems, and the coupled ABX and ABXq systems. We show that if the natural relaxation rates of nuclei in exchanging sites are identical, then in general effects of both exchange and spin coupling can be removed by pulsing sufficiently rapidly. Modulation of the echo train by spin coupling is removed by rapid exchange between the coupled sites. The Carr—Purcell experiment on an equivalent set of spins affords effective decoupling from another equivalent set unaffected by the pulses, in the absence of exchange between the sets. An analogy is drawn between chemical exchange and quadrupolar relaxation, and a method is outlined which in principle should provide an indirect determination of quadrupolar or other fast relaxation rates. We illustrate the applicability of the theory to the calculation of line shapes in steady-state spectra by treating the exchanging coupled AB system in detail, and show that in the fast exchange limit the theory predicts weak broad resonances not previously described.

The amplitude of the nuclear induction signal is calculated for pulsed nuclear magnetic resonance experiments for systems in which chemical exchange takes place between two sites of different Larmor frequency. General expressions are given, within the scope of first order rate theory for one and two pulse experiments as well as multiple pulse (Carr—Purcell) sequences. It is proven that for a Carr—Purcell sequence the decay is always expressible as a sum of two exponentials. The two time constants are derived for the general case and some special cases are discussed in detail. The theory makes possible the measurement of exchange rate constants using spin-echo techniques without the use of computers and seems relatively easy to generalize to more complicated systems.

The IgF n.1n.r. spectra of IF?, ReF7, IOFb! ReOF5, and have been recorded. 1F;gives rise to a broad symmetric doublet with a spl~tting of 4 100 f 300 c.p.s. Interpreted as part~ally collapsed and unresolved 19F-1nI coupling. ReF7 gives a single line of width 65 c.p.s. a t low field. These spectra give no information about the molecular structure other than that there is a fast intramolecular process causing magnetic equivalence. IOFs and ReOFb both exhibit a quintet/doublet spectrum indicative of locked C4, symmetry. The multiplet splittings in IOFs are not resolved in pure IOF5 or in IOF5/IF7 mixtures, but can be seen for IOF5 in solution in IF5. The doublet lies t o higher field than the quintet in IOF5, but the reverse occurs in ReOF5. In the paramagnetic molecule OsOFs only the resonance of the unique axial fluorine IS observed, indicative of localization of the non-bonding d-electron in the equitorial plane of the nlolecule.The recent controversy (1,2) over the nlolecular structure of IF7 and the demonstration in these laboratories by infrared studies (3) that much of the earlier work on IF7 was carried out on sa~nplescontaminated with IOFS have made a reinvestigation of the structure of these and related colnpounds desirable. We set out to characterize these molecules by their 19F (n.m.r.) spectra. At the onset of our investigation of the heptafluoride the only previously reported n.m.r. study was fro111 the early work of Gutowsky and Hoffman (4), who, using material of uncertain purity, reported a nlultiple spectrunl which they ascribed to nonequivalent fluorine nuclei. Preparation of the ConzpoundsThe fluorides or oxyfluorides were prepared in all-welded inonel cans provided with Hoke A431 valves joined to them by monel solder tube fittings. The cans were joined t o the fluorine line and the metal vacuum line by way of monel Swagelock fittings provided with Teflon front ferrules and nylon back ferrules.Iodine pentafluoride was made by fluorinating iodine, the reaction being moderated by immersion of the can in cold water. The more volatile fluorides were removed by vacuum distillation.Iodine heptafluoride was best made by fluorinating the distilled pentafluoride with a twofold excess of fluorine a t 200° for a n hour or so. Excess fluorine was removed under vacuum with the can held a t -19G0. Heptafluoride prepared in this way was shown by infrared spectroscopy to be oxyfluoride free.Iodine oxide pentafluoride was most conveniently prepared by the addition of a quantity of water, calculated to satisfy the eq. IF; f H2O -+ IOFG f 2HF, t o the heptafluoride which was measured out tensimetrically, then condensed in a ICelF tube. This tube was joined to the vacuuln system by a flare fitting. Hydrogen fluoride was removed by keeping the products of the reaction over sodium fluoride in a ~nonel can. Indeed, hydrogen fluoride was conveniently removed from all of the fluorides in this way.Rlreniunz oxide pentafluoride was prepared by the fluorination of rhenium dioxide as previously described (5), the r...

No abstract

THE synthesis of OCF,-salts by Willis and his co-workers1 suggested the probability of making the isoelectronic molecule, nitrogen oxide trifluoride.