Radiation damping generates surprising new features in two-dimensional and spin–echo experiments. The theoretical interpretation of the unexpected results has led to some controversy. The purpose of the present paper is to show, by computer simulations, that the experimental results can be completely understood by describing radiation damping through a classical magnetic field caused by the current induced in the sample coil (resonantly enhanced by the high quality factor of the tuned circuit), and acting back on the spins. We examine a variety of two-dimensional and ‘‘multiple quantum’’ experiments in homogeneous fields, and spin-echo experiments in a fixed field gradient (with and without molecular diffusion).

The standard theory of NMR relaxation in liquids ͑with molecular motion described as a classical Brownian motion, and including intermolecular spin-spin couplings͒ is re-examined, taking great care not to drop significant contributions from the dipolar coupling between distant molecules. This results in ''modified Bloch-Redfield equations'' for the spins in a single molecule, valid at all spin temperatures, which contain both the usual relaxation terms and a coupling of each spin with a classical average dipolar field. Delicate issues raised in this derivation, like the neglect of quantum correlations between spins on different molecules at ͑repeated͒ initial times, are discussed with the help of exact calculations ͑for all spin temperatures͒ performed on a simplified model which includes equal couplings between all N spins of a system. The same model is used to compare the merits of different forms of ''high temperature'' approximation. We also propose an iterative scheme for solving the ''modified Bloch-Redfield equations,'' in which the starting point is the well understood solution of the problem without dipolar field. Finally, a short discussion is given of the relation between ''quantum correlations'' and ''quantum coherences'' in the perspective of multiple-pulse and multiple-quantum experiments. These two notions are very simply related in the strict first approximation of weak order, and have often not been clearly distinguished. However, in the second order approximation, which is required whenever dipolar field effects manifest themselves, unrelated spins may display observable coherences although they are not coupled ͑hence not correlated͒.

No abstract

No abstract

No abstract