A brief overview is given of many topics that are covered later, followed by a detailed plan of the book. The concern is with interactions that take place between molecular dipoles in an equilibrium gas when probed by an externally sourced electromagnetic wave train. This will lead to the appearance of otherwise sharp spectral lines that may be broadened in various ways. After a brief mention of the early ideas of Lorentz and Weisskopf, the discussion moves to the real starting point for this book, which is the idea that the line shape will be determined by the fluctuating response of the active dipole to molecular collisions. Three broadening effects are distinguished. Firstly, an elastic collision at the radiating molecule may cause a sudden change in the phase of the wave train. Secondly, where an elastic collision exerts a torque on the radiator, there may be an elastic reorientation and a sudden change in the wave train amplitude. Thirdly, an inelastic collision may lead to a sudden change in the frequency of the wave train, and, if these collisions are frequent enough, there may also be interference, or coupling, between the lines as they are broadened.

A sample of gas, originally treated as a single quantum system, is now described in terms of its molecular constituents, starting with the case of a single radiating molecule in an equilibrium bath of perturbers. First, the isolated radiator is considered, as if the bath had been deactivated, allowing a discussion of how its internal energy and angular momentum may change when, in the presence of an electromagnetic field , a radiant transition takes place, and of how the transition amplitude may be reduced under the Wigner–Eckart theorem. Then, the interaction between radiator and bath is reinstated, but the initial correlations between the two are neglected, so that a separate average over the bath may be taken. There is then an examination of various approximations that may be of use elsewhere. These are the restrictions to collisions that are binary in nature, the possibility that a collision may be said to follow a classical trajectory, and the validity of treating it under the impact approximation, which carries a restriction to the core region of a spectral line, but offers a great simplification when collisions may be regarded as very brief, well-separated events.

The subject here is the absorption coefficient, expressing the net power loss from the field over a unit path. At its heart is the line shape, which may be identified with the power spectral density function for fluctuations of the active dipole in the presence of an equilibrium bath of perturbers, and, as such, should satisfy the fluctuation–dissipation theorem. The more general properties of the absorption coefficient, which must reflect this balance, are first examined in some detail, particularly for the Van Vleck–Huber form. It is then shown that this, when expanded as a sum over individual lines, may be folded into more compact expressions. Outside the line core, these expressions must incorporate the fluctuation–dissipation theorem, and special attention is given to distinguish this case and that of the core itself, where it is of no consequence. Even the very general Fano theory does not, as it stands, satisfy the theorem, and can be used for the far-wing line shape only if these expressions are modified. Finally, some account is given of how they may be used with a molecular line database, and how a calculation of radiative transfer might proceed in the simplest of cases.

Starting from the very general Fano theory of pressure broadening, ways are sought to express the shape of a band of lines in a form that is more amenable to calculation. Initially, the far-wing is considered, and care is taken to ensure, by an adjustment, that the fluctuation–dissipation theorem is satisfied, despite Fano’s neglect of the initial correlations between the states of the radiator and the bath of its perturbers. The far-wing also requires, in a Fourier sense, the use of a very fine time scale, which allows the approach taken by Rosenkranz and Ma & Tipping, described first, to adopt the quasi-static approximation. In obtaining the overall line shape, the average over collisions may then be run across an ensemble of essentially static binary configurations. In the line core, the initial correlations may be ignored anyway, and, because a much coarser time scale is appropriate, the impact approximation may be invoked. Here, Fano’s theory is shown to reduce to that of Baranger, yielding expressions for fixed line shifts and widths, and allowing, through a perturbative approximation due to Rosenkranz, a simple expression to be derived to take account of line coupling.

The focus here is on the approach taken by Anderson, which extends previous work by including the possibility that collisions will cause transitions in the radiator. Anderson confines himself to spectral lines that may be considered isolated from one another, and will, therefore, be broadened independently, and the start point is the correlation function of the radiatively active dipole, a quantum mechanical average formed from the states and operators of the gas system. This is treated as an ensemble average, in line with later chapters, and Anderson’s use of a time average is relegated to an appendix. However, the two approaches eventually converge, and both lead to a concern for the average effect on the lines as the radiator encounters an ensemble of single binary collisions on classical trajectories. Under the impact approximation, the correlation function may be greatly simplified, and expressions arise for the shift and width of a spectral line in terms of an optical cross-section that may be approached through a low order perturbative approximation. Within this, contributions due to phase shifts, elastic reorientations and inelastic transfers may all be distinguished.

The Fano theory, described here, does not adopt the impact approximation, and is not confined to the line core. The main concession, implicit in an impact theory like that of Baranger or Anderson, is the neglect of initial correlations between the states of radiator and bath, allowing a separate average to be taken over the bath variables at the initial time. For Fano, the correlation function describes the linear response to the driving field and is governed dynamically by the Liouville operator, which, in Baranger’s line space, has a role similar to that of the Hamiltonian in the original state space. The response, in its Fourier transform, provides the line shape, and this is governed here by a relaxation operator that looks, formally, like a transition operator in quantum scattering. This is a very general theory that will, nevertheless, reduce to that of Baranger as soon as the impact approximation is imposed. Although the neglect of initial correlations will invalidate the fluctuation–dissipation theorem, this will only affect the line far-wing, where, unless remedied, it will cause an imbalance between the radiative processes induced by the field.

We discuss here the choice of solid compounds and materials which best suit various types of applications, focusing mainly on the polarized targets. These materials include hydrogen-rich glassy hydrocarbons and simple cubic crystalline ammonia and lithium hydrides. The glassy hydrocarbons can doped by dissolved stable free radicals, while crystalline materials are doped by radiolytic paramagnetic radicals. The leading application of DNP up till now has been the scattering experiments in high-energy and nuclear physics. Other applications include measurements of slow neutron cross-sections, molecular physics using slow neutrons, nuclear magnetism and other solid-state physics experiments, and spin filters. The use of polarized solids in fusion and in magnetic resonance imaging has also been discussed. The material choice evidently depends strongly not only on the application but also on the goal of the experiment or process which is considered. More recently DNP has been used for the signal enhancement in NMR studies of complex chemical and biochemical molecules. In this context DNP and other enhancement techniques are called by the term “hyperpolarization”.

The principles of the continuous-wave (CW) NMR techniques is reviewed, as applied to the measurement of the nuclear polarization in polarized targets. The circuit theory of the series-tuned Q-meter is then described in detail in view of calculating precisely the CW NMR absorption signal and its integral, the signal-to-noise ratio, the probe coupling and sampling, and the signal saturation. Optimization of the series-tuned Q-meter circuit is discussed on the basis of the circuit theory. Improved Q-meter circuits will be reviewed, including the capacitively coupled Q-meter, the crossed-coil NMR circuit, and the introduction of quadrature mixer that enables the measurement of the real and imaginary parts of the RF signal simultaneously. Calibration and measurement of very small NMR signals then follows. We also treat the signal-to-noise issues, the electromagnetic interferences, and the NMR circuit drift issues.

We shall first outline the types of interactions of spins, which are most important for solid polarized targets: the magnetic dipole interaction, the quadrupole interaction, the spin-orbit interaction and the hyperfine interaction. Other direct and indirect spin interactions are described: these give rise to the chemical shift, the Knight shift, molecular spin isomers and to the exchange interaction of electron pairs. These, and in particular the dipolar interaction, are then used in the discussion of the magnetic resonance phenomena, such as the resonance line shape and saturation. The magnetic resonance absorption and the transverse susceptibility are discussed starting from the first principles, and Provotorov's equations are derived. The relaxation of spins, which is phenomenologically introduced already for the saturation, is then overviewed in greater depth, before closing with sudden and adiabatic changes of spin systems in the rotating frame.

Microwave sources and waveguide components are commercially available and therefore we shall limit ourselves to describing their physical principles and main limitations in the polarized target applications. The propagation modes and the complex propagation constants in rectangular and round guideas are derived and described. Simple waveguide components and circuits are discussed in view of control and optimization of the power, frequency and modulation for DNP. The main design principles and criteria are presented for iron core magnets and their cobalt–iron pole pieces. For superconducting solenoid and dipole magnets, the focus is in the winding accuracy. The control and protection of large superconducting magnets is also briefly discussed.