Since the earliest days of Nuclear Magnetic Resonance (NMR), it has been clear (1-10) that relaxation mechanisms were to play an important role in all its applications. In fact, even the first detection of an NMR signal (11-12) was delayed (12-13) by several years because the chosen compounds had, unfortunately, excessively long relaxation times. From the pioneering BPP (Bloembergen-Purcell-Pound) formula (7), published in 1948 itself, it became immediately evident that, qualitatively speaking, • NMR relaxation times, particularly the longitudinal ones, depend through the Larmor frequency on the magnetic field induction B0. • The relaxation mechanisms require some kind of nuclear interaction subject to stochastic fluctuations, typically due to random molecular motions. • The most pronounced relaxation phenomena (in terms of field dependence) were to be expected at relatively low fields where low-frequency molecular motions can have a very large impact on the longitudinal relaxation times T1. The dependence of T1(B) on the field B was soon nicknamed as the T1 dispersion curve or, more recently, as the Nuclear Magnetic Relaxation Dispersion (NMRD) profile. The first experimental curve of this type (Fig. 1) was published in 1950 by Ramsey and Pound (15,16). Many more such curves were measured in subsequent years, some of which were reported by Abragam (17). When Abragam's work was published it was already quite clear that the dispersion curves could become a valid tool for the study of molecular dynamics, thus laying down the foundation for variable field NMR relaxometry. In principle, the dispersion curves are potentially powerful tools to discriminate between various molecular dynamics models. The development of this branch of NMR, however, has been quite slow compared to the explosive progress of NMR spectroscopy and, later on, NMR imaging. There are many reasons for this slow start, the most obvious ones being: • Complexity of NMR relaxation theories (1,7-9,18-48). • Lack and/or complexity of molecular dynamics models. • Practical difficulties inherent in measuring the dispersion curves. Since this is essentially an engineering chapter, we shall dwell only on the last point. From the BPP formula, it was already qualitatively clear that, in order to become efficient and useful tools, the dispersion curves must extend over a wide interval of relaxation field values (preferably several orders of magnitude). Achieving this goal using the traditional, fixed-field approach was almost impossible. One can, of course, use an electromagnet and a broad-band NMR console and, re-tuning the system at every point of the measured profile, carry out a conventional relaxation time measurement at many field values. Apart from being painfully slow, however, such an approach is limited to at best one decade of rather high field values. At fields corresponding to less than about 1 MHz of the 1H Larmor frequency, the signal excitation and detection technologies in fact change too much to use the same type of instrument and, in addition, the signal becomes often too weak to be detected. Limited fixed-field, traditional relaxation measurements at very low fields, including the Earth field, were of course carried out (49) using specially built NMR systems. Such measurements confirmed the general tendency of relaxation times to be more "discriminating" at low fields than at high fields. The fact has been used even to produce a medical low-field NMR system capable of diagnosing particular fetal pathologies by means of in vivo measurements of the longitudinal relaxation time of the amniotic liquid. Even such systems, however, were limited to quite a narrow relaxation field interval. It soon became clear that in order to cover comfortably a wide range of relaxation field values, one had to use an excitation/detection assembly operating at some fixed field and, during the relaxation periods of a relaxation measurement NMR sequence, subject the sample to another, easily variable field. One rather obvious solution for such an arrangement was to combine a fixed field, conventional NMR relaxometer operating at a high field with an auxiliary variable electromagnet and, during the sequence, mechanically shuttle the sample between the two magnetic fields. For almost three decades, many T1 dispersion curves (including the first one shown in Fig. 1) were actually measured by moving the sample manually from one magnet to the other. Quite soon, however, mechanical devices (50-64) were developed to achieve the task, some of which were quite sophisticated. Since, during an actual measurement, the shuttling process is repeated many times in a cyclic manner, the technique has been named field-cycling (FC) NMR relaxometry, a term which underlines the fact that it is the magnetic field variation that matters and not the manner in which it is achieved. The main drawback of mechanical shuttling consists in the relatively long time needed to move the sample physically between the two fields. Even with the best devices, there are severe limits on the maximum acceleration/deceleration, dictated by the mechanical stability of the sample transport system as well as of the sample. Consequently, relaxation times much shorter than 100 ms (R1 above 10) are almost impossible to measure, ruling out an enormous segment of potentially interesting applications. On the other hand, mechanical shuttling systems had been successfully combined with high-field, high-resolution NMR spectroscopy (64), a feature which is still quite unique. A new approach in the early 1970s consisted in keeping the sample fixed while the field, produced by an air-core electromagnet, is switched between different field values. This approach, named fast field-cycling (FFC) NMR relaxometry, explored primarily by Redfield (35,75) Noack (61,65,67,77) Koenig (58,66,78) and Kimmich (61,73,76,77,94) has the potential of handling much faster relaxing samples (the current upper limit of manageable R1 is between 1000 and 10000, depending upon the shape of the NMRD profile). It requires novel type of magnets and power supplies, the development of which is still in progress. The advent of the FFC instruments has opened a number of important application areas (molecular dynamics of liquid crystals, paramagnetic contrast MRI agents, proteins, polymers, etc.) and has thus provided a powerful impulse for further development of variable-field NMR relaxometry. Since 1996, Stelar entered the field and, building on the Noack-Schweikert technology (67), started producing the first commercial FFC NMR relaxometers. The availability of such instruments has further enhanced the drive towards new applications, apart from confirming the enormous potential of the technique as a primary tool for the study of molecular dynamics of even quite complex systems. In what follows, we wish to describe the most important technical aspects of FFC NMR relaxometry, including both the required special hardware (magnet, power supply, etc.) and the measurement methodology (data acquisition sequences and, to some extent, the subsequent data evaluation). Naturally, the description is based primarily on our own experience which has not yet been described in detail elsewhere. © 2005 Elsevier Inc. All rights reserved.
