Laboratory Test of the Isotropy of Light Propagation at the ${10}^{-17}$ Level

We report on the results of a strongly improved test of local Lorentz invariance, consisting of a search for an anisotropy of the resonance frequencies of electromagnetic cavities. The apparatus comprises two orthogonal standing-wave optical cavities interrogated by a laser, which were rotated approximately 175 000 times over the duration of 13 months. The measurements are interpreted as a search for an anisotropy of the speed of light, within the Robertson-Mansouri-Sexl (RMS) and the standard model extension (SME) photon sector test theories. We find no evidence for an isotropy violation at a $1\sigma$ uncertainty level of 0.6 parts in ${10}^{17}$ (RMS) and 2 parts in ${10}^{17}$ for seven of eight coefficients of the SME.

Figure 1

(a) schematic of the optical setup. The ULE cavity block is temperature stabilized inside a vacuum chamber (textured area). The wave ${\nu }_3$ is used for prestabilizing the frequency of the laser to a ${\mathrm{TEM}}_{01}$ mode of one cavity. PD1, PD2, PD3 are detectors used to generate error signals for controlling the frequencies of AOM1, AOM2 and the laser itself, respectively. PD4 monitors the transmitted power of wave ${\nu }_3$. PD, photodetector; AOM, acousto-optic frequency shifter and modulator; EOM, electrooptic phase modulator; PBS, polarization beam splitter (BS). (b) Side cutaway view of the whole apparatus. A, air cushion or bearing; B, rotary motor; C, voice coil actuators; D, two-axis tilt sensor; E, air spring system; TEC, thermoelectric cooler elements for temperature regulation.

Figure 2

Histograms of the measured modulation amplitudes (all data). For comparison, the dots indicate the histogram values of a Gaussian distribution with the same mean and variance as the corresponding data set.

Figure 3

Isotropy violation signals. Points: mean values of the modulation amplitudes measured within 0.5 h intervals during each sidereal day (all data). The error bars indicate the statistical standard errors of the means. Full lines (blue): range of values that the RMS violation signals (not averaged) for the fitted value $\delta -\beta +\frac{1}{2}$ of Eq. (2) and a systematic offset on $B$ take on during 1 yr. The two full lines lie very close since Earth’s orbital velocity is small compared to the solar system velocity. Dashed lines (red): same for the SME violation signals (not averaged) given by the fitted values of Eq. (3).