Coriolis effect is an important error source in the weak equivalence principle (WEP) test using atom interferometer. In this paper, the problem of Coriolis error in WEP test is studied theoretically and experimentally. In theoretical simulation, the Coriolis effect is analyzed by establishing an error model. The measurement errors of Eötvös coefficient (η) in WEP test related to experimental parameters, such as horizontal-velocity difference and horizontal-position difference of atomic clouds, horizontal-position difference of detectors, and rotation compensation of Raman laser’s mirror are calculated. In experimental investigation, the position difference between 85Rb and 87Rb atomic clouds is reduced to 0.1 mm by optimizing the experimental parameters, an alternating detection method is used to suppress the error caused by detection position difference, thus the Coriolis error related to the atomic clouds and detectors is reduced to 1.1× 10−9. This Coriolis error is further corrected by compensating the rotation of Raman laser’s mirror, and the total uncertainty of η measurement related to the Coriolis effect is reduced as δη = 4.4 × 10−11.

The equivalence principle (EP) is a basic assumption of the general relativity. The quantum test of the equivalence principle with atoms is an important way to examine the applicable scope of the current physical framework so as to discover new physics. Recently, we extended the traditional pure mass or energy tests of the equivalence principle to the joint test of mass–energy by atom interferometry (Zhou et al.,Phys.Rev.A 104,022822). The violation parameter of mass is constrained to η0 = (−0.8 ± 1.4) × 10–10 and that of internal energy to ηE = (0.0 ± 0.4) × 10–10 per reduced energy ratio. Here, we first briefly outline the joint test idea and experimental results, and then, we analyze and discuss how to improve the test accuracy. Finally, we report the latest experimental progress toward a high-precision mass–energy test of the equivalence principle. We realize atom interference fringes of 2T = 2.6 s in the 10-m long-baseline atom interferometer. This free evolution time T, to the best of our knowledge, is the longest duration realized in the laboratory, and the corresponding resolution of gravity measurement is 4.5 × 10−11 g per shot.

We demonstrate a novel laser frequency shift scheme using a 12-pass 350-MHz acousto-optic modulator (AOM). This AOM system shows better performance compared to ordinary acousto-optic modulation schemes. The frequency of the incident laser beam is shifted by 4.2 GHz with the total diffraction efficiency as high as 11%, and the maximum frequency shift is 5 GHz. Combining the ±1st order diffraction, laser signals with up to 10 GHz frequency difference can be obtained, which fulfill most frequency shift requirements of laser cooling and coherent manipulation experiments with alkali metal atoms.