Picometer control of a levitating milligram gravity sensor

Dennis G. Uitenbroek, Jurriaan Langendorff, Tjerk H. Oosterkamp · 2026

Read Picometer control of a levitating milligram gravity sensor by Dennis G. Uitenbroek, Jurriaan Langendorff, Tjerk H. Oosterkamp (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-28479.txt.

Introduction

Dennis Uitenbroek, Jurriaan Langendorff, and Tjerk Oosterkamp cool a floating milligram gravity sensor with linear feedback to great stillness. A magnet floats in a superconducting trap, its motion read by a coil linked to a SQUID and damped through a piezo actuator. Two side-to-side modes reach below two picometers of motion and below ten millikelvin at once. The result moves quiet mounting and sensing toward future quantum ground-state cooling of levitated particles.

Catalog record

CREATORDennis G. Uitenbroek, Jurriaan Langendorff, Tjerk H. Oosterkamp
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE38959
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.28479

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Picometer control of a levitating milligram gravity sensor Dennis G. Uitenbroek, Jurriaan Langendorff, Tjerk H. Oosterkamp Abstract Due to their exceptional isolation from the environment, magnetically levitated particles are explored as extremely sensitive mechanical sensors. For future gravity experiments on quantum superpositions, such systems need to be cooled close to their ground state. To demonstrate the combination of state of the art vibration isolation, milligram levitated high Q mechanical resonators and position detection with low noise, we present linear feedback cooling of a magnetically levitated gravity sensor to below 2 picometer amplitude and below 10 millikelvin mode temperature for two translational modes (the x- and y-mode) simultaneously. The sensor is a levitating permanent magnet in a type I superconducting trap, where its six resonance frequencies are measured with a superconducting coil coupled to a DC SQUID. This signal is measured with a lock-in amplifier and a feedback signal is sent to a piezoelectric actuator, allowing the cooling of resonant modes at 50.6 50.6\text{\,} and 68.0 Hz 68.0\text{\,}\mathrm{H}\mathrm{z} simultaneously. These two translational modes have Q factors of 3.8 ⋅ 10 6 3.8\text{\cdot}{10}^{6}\text{\,} and 5.5 ⋅ 10 6 5.5\text{\cdot}{10}^{6}\text{\,} respectively. The experiment is mounted inside a dry dilution refrigerator where it is vibrationally attenuated with 110 - 130 dB 110130\text{\,}\mathrm{d}\mathrm{B} at these frequencies. In this work, we discuss future improvements on the setup which may enable quantum ground state cooling on a magnetically levitated particle, that has previously been shown to be a gravitational sensor. \mysection Introduction Large mass quantum experiments are the key ingredient in closing the gap between microscopic quantum systems and macroscopic general relativity. By pushing the quantum effects to larger and heavier scales, an experiment combining these two theories could be enabled. Observing quantum mechanical effects in a mechanical system requires a crucial first step: cooling a particle close to the mechanical quantum ground state. This has been a goal within the physics community on various platforms using different cooling methods. Additionally, one would like to achieve this quantum regime in a mass regime where gravity becomes relevant. In early work, micromechanical membranes were shown to be cooled to the quantum ground state 17 ; 23 ; 3 ; 20 ; 10 . These type of systems will inevitably suffer from clamping losses, due to the suspended nature of these experiments. In an effort to avoid clamping losses, levitated systems can be studied. Trapped ions were the first to be cooled to the mechanical groundstate, using laser sideband cooling 5 ; 16 ; 14 . Although impressive results are accomplished in this field, these type of experiments are limited to very few atoms in scale. In optical levitation, groundstate cooling has been achieved 4 ; 22 ; 15 ; 19 since a couple of years. These experiments are performed at relatively high frequencies ( 10\text{\,}\mathrm{k}\mathrm{Hz} - 1\text{\,}\mathrm{M}\mathrm{Hz} ) compared to magnetic levitation. At these frequencies mechanical vibrations are less present and can be readily attenuated. This is a very promising platform for pushing the limits of quantum mechanical effects. However, these particles have a size around 100\text{\,}\mathrm{n}\mathrm{m} and a mass in the order of femtograms. This is already a large step in macroscopicity compared to ions, but still far from a macroscopic particle that would generate a measurable amount of gravity. A different levitation platform, magnetic levitation, makes use of magnetic forces. This platform is not limited in mass due to the magnetic forces and there is no internal heating of the particle as in the above-mentioned levitation systems. With the larger mass the frequencies will decrease (compared to optical levitation). At these lower frequencies dampening the mechanical vibrations is more challenging. Different types of magnetic levitation exist, we will only discuss superconducting levitation here, superconducting levitation can be divided in active 28 ; 12 ; 13 ; 9 ; 11 ; 21 ; 18 ; 7 and passive systems 24 ; 29 ; 31 ; 25 . Active systems send a current through coils that levitate a piece of superconducting material by using the perfect diamagnetic properties of the superconductor. The passive systems use the same perfect diamagnetic property due to the Meissner effect and let the superconductor expel the magnetic field, coming from a permanent magnet. Both mechanisms are, in principle, scalable to sizes greater than 1\text{\,}\mathrm{m}\mathrm{m} . The advantage of a passive system is the absence of noise sources from the active components, e.g. the current noise in the coils. Passive magnetic levitation with a submillimeter particle of 0.43\text{\,}\mathrm{m}\mathrm{g} has already been demonstrated in a gravitational experiment 6 . In this letter, a version of this setup with state-of-the-art vibration isolation, a massive levitated particle with high Q factors, and low noise readout are all together combined to reduce the motion of the particle to below two picometer of RMS amplitude. This achieved motion is a benchmark for the performance of the system, this can only be achieved if all necessary experimental conditions are fulfilled simultaneously. Performing ground state cooling on these type of systems is possible in theory, but experimentally very challenging. The zero point motion scales with \sim 1/\sqrt{m\omega} . In magnetic levitation the relatively large masses in combination with low frequencies result in a small zero point motion, 0.7\text{\,}\mathrm{f}\mathrm{m} for the setup used here. In this experiment a phonon number in the order of {10}^{6}\text{\,} is reached. The theoretical minimum phonon limit with current detection and force noise is in the order of {10}^{5}\text{\,} phonons. With improvements on vibration isolation, Q factors, detection coupling and SQUID noise, we think it would be possible to cool this system to the ground state. Since this setup has already been used for gravitational measurements 6 , combining this system with the feedback cooling presented in this paper, this experiment demonstrates to be a promising platform towards combining both gravitational interaction and quantum mechanical effects in a single experiment. \mysection Experimental setup The experiment consists of a permanent magnet levitating in a superconducting trap by Meissner repulsion. The motion of the magnet is measured using a Superconducting QUantum Interference Device (SQUID) measured by a lock-in amplifier. The lock-in then calculates a feedback signal and sends this to a piezoelectric actuator, schematically shown in Fig. 1 a. The permanent levitating magnet is composed of three 0.25\text{\,}\mathrm{m}\mathrm{m} by 0.25\text{\,}\mathrm{m}\mathrm{m} by 0.25\text{\,}\mathrm{m}\mathrm{m} cubic neodymium iron boron parylene coated magnets stacked and glued together, using Stycast. A glass sphere with a diameter of 0.20\text{\,}\mathrm{m}\mathrm{m} is attached to this stack of magnets, again using Stycast, see Fig. 1 b. The glass sphere is placed off center to break symmetry. Typical permanent remnant magnetisation of these magnets is 1.4\text{\,}\mathrm{T} , although it is believed that at low temperatures this is reduced by 2025\text{\,}\% 26 . The estimated mass of this composed magnet and glass sphere is 0.35\text{\,}\mathrm{m}\mathrm{g} . The magnet levitates in a type I superconducting trap made of tantalum, shown in Fig. 1 c. This trap has an elliptical shape, 3.0\text{\,}\mathrm{m}\mathrm{m} by 4.0\text{\,}\mathrm{m}\mathrm{m} as the minor and major axes. The pick-up coil is placed at a height of 4.0\text{\,}\mathrm{m}\mathrm{m} from the bottom of the trap. The calculated levitation height of the magnet is 1.96\text{\,}\mathrm{m}\mathrm{m} 29 . The pick-up coil is able to detect the motion of the magnet, via the induced

Read the complete text