For nuclear magnetic resonance (NMR) spectroscopy and tomography radio signals sent out by magnetically active atomic nuclei in a magnetic field are analyzed in order to obtain chemical or physiological information. In today's routinely applied NMR methods radio frequency pulses effect the excitation of these signals.
In the completed project we investigated, how chemical information can be obtained from a sample in a magnetic field without excitation, only from analysis of the undisturbed radio frequency noise (NMR-noise) of a sample in a magnetic field.
First the methods for acquisition of the noise data were optimized. Then efficient algorithms for analysis of the NMR-noise were implemented.
We could show that from noise data the hydrogen NMR-spectra of chemical compounds in equilibrium conditions can be determined. The phenomenon of spin-noise itself was thoroughly investigated. We developed a model that describes NMR noise as two components: the pure spin noise and the absorbed circuit noise. This model allows the explanation of the complex behaviour of NMR-noise during NMR-spectroscopic and imaging experiments. Based on the observation of NMR noise we developed a new method for tuning of NMR-probes, which allows one to increase sensitivity by up to 40%. Applications of this procedure were demonstrated both on protein solutions and solid samples of small molecules.
Potential applications of the results lie in the development of NMR methodology (spectroscopy and imaging) of very small sample amounts (less that 107 molecules).
From a theoretical point of view the project results have yielded fundamental insights into the NMR phenomenon. These have implications for the development of new NMR techniques like dynamic nuclear polarization, ultra-low field and ex-situ NMR.