Cryogenic pressure sensing with an ultrafast Meissner-levitated microrotor

TL;DR

The paper investigates cryogenic gas-pressure sensing using a Meissner-levitated microrotor. It demonstrates wide-range sensing from \\(P \\sim 10^{-3} \\text{to} \\ 10^{-8} \\text{mbar}\\ by measuring the spin-down rate \\gamma of a micromagnet levitated in a superconducting trap at \\(T = 4.2 \\text{K}, with \\gamma matching gas-damping predictions \\(\\gamma \\approx \\frac{10}{\\pi} \\frac{P}{\\rho \\bar{v} R}. The rotor spins up to \\(f_s = 2.3 \\text{MHz} (138 \\text{Mrpm}) with \\(Q \\approx \\pi f / \\gamma \\approx 10^{13}$, and a low-frequency precession follows \\(f_l \\sim f_x^2 / f_s) with \\(f_x \\approx 240 \\text{Hz}, validating the model. This cryogenic Spinning-Rotor Gauge provides an ultralow-torque-noise, absolute pressure standard for cryostats and holds promise for quantum sensing, gravitational tests, and fundamental physics experiments at millikelvin temperatures.

Abstract

Magnetically levitated spinning rotors are key elements in important technologies such as navigation by gyroscopes, energy storage by flywheels, ultra-high vacuum generation by turbomolecular pumps, and pressure sensing for process control. However, mechanical rotors are typically macroscopic and limited to room temperature and low rotation frequencies. In particular, sensing pressure at low temperatures remains a technological challenge, while emerging quantum technologies demand a precise evaluation of pressure conditions at low temperatures to cope with quantum-spoiling decoherence. To close this gap, we demonstrate wide range pressure sensing by a spinning rotor based on a micromagnet levitated by the Meissner effect at 4.2 Kelvin. We achieve rotational speeds of up to 138 million rotations per minute, resulting in very high effective quality factors, outperforming current platforms. Beside sensing applications, we envision the use of levitated rotors for probing fundamental science including quantum mechanics and gravity, enabled by ultralow torque noise.

Figures (5)

• Figure 1: A: General scheme of the experimental setup, including a micromagnet levitated in a superconducting lead trap, a detection system based on a dc SQUID and a pick-up coil (PC), and a feedback circuit based on superconducting driving coils (DC). Spinning of the magnet on the horizontal plane is controlled by a synchronous driving technique. Variable gain $G$ and phase $\phi$ allow controlling magnitude and sign of the torque. B: Conventions on the reference frame and angles. C: Simplified three-dimensional sketch of the main elements of the setup.
• Figure 2: A: Example of single spectrum acquired during a free spin-down. We estimate the instantaneous frequency as the frequency corresponding to the maximum of the peak in the spectrum. B: Example of full spin-down curve at the pressure $P_g=1.12 \times 10^{-3}$ mbar. For each point the frequency is estimated from the spectrum as in Fig. A. The red line shows an exponential fit of the spin-down regime. The abrupt interruption of the exponential spin-down at $t\sim 400$ s corresponds to the transition from free rotational motion to librational trapping, with subsequent relaxation of the frequency towards the small libration value.
• Figure 3: A: Frequency decay rate as function of the pressure $P_g$ measured by the room temperature gauge. B: Zoomed-in plot of the low pressure region, with the best linear fit to the data.
• Figure 4: Spin-down curve measured at a frequency $f\sim 2.01$ MHz, at the lowest effective cryogenic pressure achieved in the experiment. The data are well-fitted by the exponential curve Eq. (\ref{['spindown']}) with $\gamma= 4.75 \times 10^{-7}$ s.
• Figure 5: A: Spectra taken from the same spin-down experiment at different times (1,2,3). The large peak at high frequency is the main spinning signal. A much weaker peak appears at low frequency, with frequency inversely related to the spinning frequency. We attribute this peak to a precessional motion. The peak between low-frequency peaks labelled 2 and 3 corresponds to the $z$ translational mode. B: Frequency of the low frequency peak $f_l$ as a function of the spinning frequency $f_s$. The red line is a fit with the equation $f_l = f_x^2 /f_s$.