New Experimental Limits on Macroscopic Forces Below 100 Microns

TL;DR

This work reports the strongest constraints to date on new macroscopic forces with Yukawa ranges from 5 to 500 μm, achieving sensitivity of about 4× gravity near 200 μm and substantially tightening limits in the 10–100 μm window. Using 1 kHz planar test masses separated by a stiff shield and measured with a capacitive readout in high vacuum, the experiment implements a Bayesian likelihood framework that marginalizes over systematic uncertainties to derive 95% CL limits on the Yukawa strength α as a function of λ. The null result, consistent with detector thermal noise, rules out large portions of parameter space predicted by string-inspired models with low-energy SUSY breaking, moduli, and dilaton scenarios, and motivates pursuit of even thinner shields and optimized geometries to push sensitivity toward shorter ranges (10–50 μm). The findings advance experimental scrutiny of submillimeter gravity and provide quantitative benchmarks for future tests of extra dimensions and scalar fields in the near-field regime.

Abstract

Results of an experimental search for new macroscopic forces with Yukawa range between 5 and 500 microns are presented. The experiment uses 1 kHz mechanical oscillators as test masses with a stiff conducting shield between them to suppress backgrounds. No signal is observed above the instrumental thermal noise after 22 hours of integration time. These results provide the strongest limits to date between 10 and 100 microns, improve on previous limits by as much as three orders of magnitude, and rule out half of the remaining parameter space for predictions of string-inspired models with low-energy supersymmetry breaking. New forces of four times gravitational strength or greater are excluded at the 95% confidence level for interaction ranges between 200 and 500 microns.

Figures (7)

• Figure 1: Central components of the apparatus
• Figure 2: Major components of the detector readout and PZT drive
• Figure 3: Data from the biased runs of 10 consecutive cycles of the data sample. Points from separate cycles are labeled with unique symbols. 1 $\sigma$ error bars are shown for one cycle. Groups of points corresponding to a particular drive frequency are circled.
• Figure 4: Distributions of data samples without shield bias. Left--hand plots: Drive (and lock-in reference) tuned to detector resonance. Right--hand plots: Drive tuned 2 Hz below detector resonance.
• Figure 5: Means of the distributions in Fig. \ref{['fig:raw']}. with channels plotted against each other to show phase. Error bars are 1 $\sigma$ standard deviations of the means. The point with the larger error bars is the on-resonance mean.