Investigating 15N@C60 using pulsed EPR and the use of new techniques in resonator production

O'Shaughnessy Gutierrez, Michael and Laird, Edward (2026) Investigating 15N@C60 using pulsed EPR and the use of new techniques in resonator production. Masters thesis, Lancaster University.

Text (2026MichaelO'ShaughnessyGutierrezMSc)
2026MichaelO_ShaughnessyGutierrezMSc.pdf - Published Version
Download (22MB)

Abstract

Improvements in timekeeping offer a wide range of benefits, and are of particular interest for global navigation satellite systems (GNSS). The accuracy and reliability of these can be dramatically improved with an improved receiver clock. A chip scale atomic clock (CSAC) could offer greatly improved performance when compared to current quartz oscillators. A CSAC is under investigation at Lancaster University, using 15N@C60, this is a Nitrogen-15 atom trapped inside a fullerene cage, and showsgreat potential for use in a CSAC; it offers potential improvements in size, weight and power in comparison to other CSAC technologies, which are very important in making a CSAC feasible to use, especially in consumer electronics. The Lancaster group’s prototype clock has room for improvement. Firstly, an investigation into potential improvements in the resonator was conducted. Litz wire and 3D printed resonators were developed and tested, showing worse Q factors than currently used solenoid resonators. A 3D printed form for making new solenoid resonators was created, and improved the creation process of new solenoids. Clock stability is parametrised by Allan deviation, which measures the rate of drift of a clock; this has not yet been measured at the clock transition. Progress has been made towards this in the form of a measurement of T2 relaxation time in 15N@C60 at 60MHz. This was found to be 3.9 ± 0.5µs and 5 ± 1µs for 2.1mT and 2.0mT transitions respectively, giving frequency linewidths of δf = 80 ± 10kHz (2.1mT) and δf = 60 ± 10kHz (2.0mT). Taking these to be the clock transition linewidths, Allan deviation estimates of: σy(τ ) = 2.2 ± 0.3 × 10−4τ1/2 for 2.1mT, and σy(τ ) = 1.6±0.3×10−4τ1/2 for 2.0mT, were calculated. Finally, an investigation into the signal-to-noise (S/N) ratio was carried out: the noise was found to be higher than theoretical levels, and did not reduce with averaging as much as expected. While the Allan deviation estimates are low, the issues with S/N ratio show clear potential for improvement in the setup, which may allow for a measurement of the line width to be made at the clock transition.