Magnetic resonance force microscopy (MRFM) is a scanning probe technique capable of producing high-resolution 3d magnetic resonance imaging (MRI) data from nanoscale sample volumes. The detection mechanism is based on the interaction between unpaired spins in the sample and a magnetic particle on a mechanical resonator, which transduces resonant spin flips into resonator motion. Several proof of concept experiments have demonstrated the detection of individual spins and subnanometre spatial resolution. However, practical implementations of the technique so far have not been able to achieve the same.
In this thesis a new design for a MRFM detector based on a carbon nanotube (CNT) nanoelectromechanical system (NEMS) is proposed. CNT resonators, due to their low mass, high compliance and high mechanical quality factor, are known to be excellent force detectors. The main challenge in fabrication of a CNT based MRFM detector is deposition of a high purity magnet on the nanotube without causing device degradation from fabrication induced defects and contamination. The proposed device design addresses this by using the substrate as shadow mask for material deposition on the nanotube. A corresponding fabrication process based on CNT stamp transfer is developed, aiming to produce ultraclean devices.
A device is successfully fabricated (skipping magnet deposition for initial characterisation) and characterised at millikelvin temperatures. It exhibits Coulomb blockade with minor disorder in the hole conduction regime and fully blocked current in electron conduction. Sixteen mechanical resonances in the MHz range with quality factors up to about 2 × 105 are identified, some of which are suspected to be parametric excitations.
Finally, first tests of a new method for the detection of nanotube motion are presented. Electromechanical mixing is used to upconvert CNT mechanical frequencies to the GHz range and the signal is detected with a near quantum-limited Josephson travelling wave parametric amplifier as primary gain stage. Initial results seem promising, with measurement sensitivity clearly exceeding that of detection via electromechanically rectified dc current. Nanotube oscillation without external drive, presumably current-driven self-oscillation, is observed. Unfortunately, more data is still required to unambiguously identify the origin of non-driven oscillation and for determining the force sensitivity of the experiment.