The atomic point contact as a detector of nanomechanical motion

I experimentally demonstrate that an atomic point contact (APC) is a sensitive detector of nanomechanical motion. With a microwave technique, I increase the measurement speed of APCs by a factor of 500. This measurement is fast enough to detect the resonant motion of nanomechanical structures at frequencies up to 150 MHz. I measure displacement with a shot-noise limited imprecision of Sx = 0.29 fm/ Hz and simultaneously observe a SF = 61 aN/ Hz backaction force. A quantum limited detector would operate at the limit imposed by the Heisenberg uncertainty principle, SxSF ≥ h; for this APC detector SxSF = 168h. Because the measurement noise is dominated by the shot noise of tunneling electrons, the non-ideality of the APC detector is likely due to a backaction force in excess of that required by quantum mechanics. Although I cannot unambiguously determine the origin of this excess backaction force, I am able to eliminate certain possible origins. For example, the observed linear dependence of the backaction force SF on APC current is inconsistent with a noisy electrostatic attraction mediated by the mutual capacitance between the APC electrodes. In contrast, a model of the backaction force that invokes a momentum impulse delivered by each tunneling electron correctly accounts for the observed scaling. However, each electron would have to deliver a momentum impulse greater than 20 times the Fermi momentum which seems implausibly large. I also observe the signs of molecular vibrations in the APC using inelastic electron tunneling spectroscopy. At the bias energy associated with vibrations I measure a resonant increase in the backaction force. This observation suggests that the excess backaction may arise from the interaction of tunneling electrons and molecular vibrations.