Spin-Lattice Relaxation Below IK

NMR experiments at very low temperatures have been carried out for a long time, generally not as a general purpose method for analyzing structure and dynamics but rather as a convenient diagnostic tool in the study of substances whose special properties at very low temperatures are of interest. These include most prominently solid hydrogen and solid and liquid helium-3. It must always have been obvious that the "routine" applications of NMR would often benefit from the use of low temperatures. The most obvious reason is the greatly enhanced sensitivity which in principle can be obtained; NMR is ordinarily spectacularly insensitive because the Zeeman levels involved are separated only by a very small energy difference fwo, yet at ordinary temperatures huio/kT is so small that the levels are almost equally populated and the excess of absorption over stimulated emission is very small. The manner in which this situation improves with lower temperatures is indicated in Figure 1, where we have chosen the particular case of protons with WO/2TT = 144 MHz as an example. Three features are evident from the figure: 1) a gain in sensitivity of 10 10(10 1O in terms of power sensitivity or averaging time) is available at the lowest temperatures. 2) A point of diminishing returns is reached at temperatures of a fewmillikelvin where hu>o ~ kT, meaning that one need not be concerned about the difficult technology for reaching temperatures lower than that; fortunately helium dilution refrigerators, capable of reaching the millikelvin regime, are now commercially available. 3) If one's goal is higher sensitivity it is foolish to stop at, say, the easily reached boiling point of helium since only a tiny fraction of the potential improvement is realized at that point.