NMR Without the Magnet or RF Coils

To image an object's interior with nuclear magnetic resonance (NMR) a magnetic field of several tesla (1 T =10,000 gauss) is usually required to polarize protons in the sample and then radio waves are used to tip the protons and to detect a weak signal as they upright themselves again. The strength of the signal depends on the size of the magnetic field and the degree of polarization, which is often only one part in 10⁵, and somewhat limits the use of NMR (including its medical application, MRI) because of the need for a bulky, expensive magnet. One way of improving things is to use laser light to produce a polarization as high as 10% in a gas of xenon atoms. The Xe atoms can then be injected into an empty space, such as lungs, and used to image their interior, which couldn't be done using conventional NMR (see Update 398). Another NMR advance has been the use of ultrasensitive SQUID detectors for picking up the magnetic fields produced by protons, greatly reducing the need for large magnets (see Update 528) but at the expense of weak signals, with a proton polarization of only one part in 10⁸.

Now, Princeton physicist Michael Romalis and co-workers, while studying whether the Xe nucleus is slightly nonspherical (equivalent to saying that the nucleus possesses a nonzero electric dipole moment, which would imply the existence of "new physics" beyond the Standard Model), have worked out a way to combine different techniques to obtain a strong NMR signal in a very weak 1 micro-tesla magnetic field. They transfer polarization from laser-polarized Xe to protons in an organic liquid and then use SQUID detectors to measure the magnetic field produced by the polarized protons. Romalis (609-258-5586) expects that this low-field NMR technique would work for any sample---whether liquid, surface, or biological tissue---with good solubility for xenon. (Heckman et al., Physical Review Letters, upcoming article; see also Princeton website)