Nanoscale Dynamic Nuclear Polarization in Force-Detected Magnetic Resonance

Abstract

Nuclear magnetic resonance (NMR) has played a pivotal role in modern science with its ability to perform non-destructive imaging and spectroscopy of various systems. Despite this, NMR has been plagued by low detection sensitivity when trying to study nanoscale ensembles of spins, primarily due to the small thermal polarization of nuclear spins. Extending the capabilities of NMR to address nanoscale sample volumes would present exciting opportunities for studying biological systems, enabling high-resolution imaging of single biomolecules and virus particles. Over the years, a great deal of techniques to improve the detection sensitivity of the measurement apparatus have been made. Force-detected magnetic resonance is one such technique, that has demonstrated the capability to detect nanoscale ensembles of spins, where it has been successfully used to achieve three dimensional images of virus particles. Nevertheless, further improvements are needed to achieve high resolution atomic scale imaging of nanoscale systems. Techniques such as dynamic nuclear polarization (DNP) have been widely implemented in traditional NMR experiments for boosting the signal, by transferring the comparatively larger polarization of electrons to surrounding nuclei . The use of DNP in force-detected magnetic resonance platforms however, has remained relatively limited though. Bringing DNP to nanoscale force-detected magnetic resonance setups would mark a significant next step in improving the detection sensitivity of nanoscale NMR experiments.

In this thesis, we discuss the implementation of DNP in a force-detected magnetic resonance experiment in order to achieve sensitivities needed to realize high resolution imaging of nanoscale spin ensembles. In these experiments, we observed a 100 fold enhancement in the proton thermal signal in a nanoscale droplet composed of trityl-OX063 radicals suspended in a sugar-water glassy matrix. We also compare the signal-to-noise ratio (SNR) boost this provides over measurements that rely upon statistical polarization, where we demonstrate a reduction in averaging time by a factor of 204. This work explores various tunable parameters to optimize the enhancement such as the proton and radical relaxation times. This work also investigates the role fast-relaxing paramagnetic defect centers from the surrounding environment play in reducing the radicals spin-lattice relaxation time, a crucial component for efficient DNP.