† Corresponding author. E-mail:
Project supported by the Strategic Priority Research Program of the Chinese Academy of Sciences (Grant No. XDB 21030200) and the National Natural Science Foundation of China (Grant No. 11904408).
The measurement performance of the atom interferometry absolute gravimeter is strongly affected by the ground vibration noise. We propose a vibration noise evaluation scheme using a Michelson laser interferometer constructed by the intrinsic Raman laser of the atomic gravimeter. Theoretical analysis shows that the vibration phase measurement accuracy is better than 4 mrad, which corresponds to about 10−10 g accuracy for a single shot gravity measurement. Compared with the commercial seismometer or accelerometer, this method is a simple, low cost, direct, and fully synchronized measurement of the vibration phase which should benefit the development of the atomic gravimeter. On the other side, limited by equivalence principle, the result of the laser interferometer is not absolute but relative vibration measurement. Triangular cap method could be used to evaluation the noise contribution of vibration, which is a different method from others and should benefit the development of the atomic gravimeter.
Since 1990s, atomic interferometers based on matter wave interference theory and laser cooling technique have been applied widely in many aspects,[1–3] such as atomic gyroscope,[4,5] atomic gravimeter,[6,7] and gravity gradiometer.[8,9] As one of the most important applications, the atomic gravimeter has reached an ultrahigh-sensitivity of 42 nm/(s2⋅Hz1/2),[10] an accuracy of 39 nm/s2 and a long-term stability of 0.5 nm/s,[11] which is compete with the state of the art classical absolute gravity measurement instruments.[12] However, it is far below the ultimate performances of atomic interference device because the noises come from the Raman laser phase and mirror vibration. The mirror vibration affects the Mach–Zehnder (MZ) atomic interferometric phase by creating an additional phase noise on the counter-propagated Raman lasers, in this way the interference fringe contrast will decrease and the measurement uncertainty will increase.[13,14] Therefore, several techniques are used to mitigate or isolate the effect of the vibration noise in atomic gravimeter. These vibration noise suppression techniques can be classified into two kinds: the first kind is reducing the absolute vibration of the mirror by putting it on a passive or active vibration isolation platform, in this way the vibration noise below 10 Hz could be reduced to a level of 10−8 g/Hz1/2,[13,15] the second kind is monitoring the vibration of the mirror with high sensitive seismometer or accelerometer, and then correcting the vibration phase with real-time compensation or post-correction techniques.[16,17] The real-time compensation technique feedbacks the monitored vibration noise power spectral densities to the Raman laser phase-locked device to compensate vibration noise while the post-correction technique compensates the vibration noise induced phase fluctuations in the final atomic interference phase. However, from the perspective of implemental technology, these methods either need to add a cumbersome vibration isolation platform or an expensive seismometer to the atomic gravimeter. From the perspective of the noise suppression results, these methods are inconsistent with the vibration of the mirror.
Benefit from the ultrahigh measurement precision, the Michelson laser interferometer (MLI) has been widely used in many areas, such as gravitational wave detection,[18] distance measurement,[9,20] and vibration detection.[21,22] In this paper, we propose a scheme combing the MLI with atomic interferometer (AI) to monitor the vibration of the Raman reflection mirror, which is essential to estimate the vibration noise induced gravity measurement uncertainty in atomic gravimeter. This MLI is constructed of the intrinsic Raman laser of the atomic gravimeter and a four-channel phase shift detector.[21,23] Thanks to the ns-level response speed of the phototube and the high accuracy of the laser frequency,[24–26] this method could achieve a higher measurement speed and accuracy than the seismometer or accelerometer. We also discuss the absolute vibration measurement problem and give a modified method based on the “triangular cap method”.[27]
The atomic interferometer gravimeter (AIG) is usually built using an atomic fountain or an atomic freefall inside a vacuum chamber device. We take the type of atomic fountain as an example. In one measurement cycle, cold atoms are prepared and launched in vertical direction firstly, and then three “π/2–π–π/2” Raman laser pulses are implemented to split–reflect–combine the cold atoms to construct the MZ interferometer in their trajectory of parabolic motion. The Raman lasers with wave vectors of k1 and k2 propagating downwards along the vertical axis of the device are reflected upwards by a mirror below the vacuum chamber, as shown in Fig.
The measurement principle of the mirror vibration using the MI of Raman lasers (MIRL) is shown in Fig.
In the detection part, we adopt the four-channel phase shift detection method[21,23] to improve the measurable accuracy and range. Four channels of coherent signals with fixed additional phase of 0, π/2, π, 3π/2 generated by the wave plates and PBSs are detected by photodetectors of PD1--PD4. The detected signals are expressed as
In the signal processing part, two quadrature signals are obtained by subtracting I1 from I3 and I2 from I4. The real-time interference phase shift δϕ (t) of the MIRL could be acquired by calculating the arc-tangent value of the two quadrature signals and distinguishing with a bidirectional counting and fringe subdivision part. Consider that the reference arm R of the MIRL usually is short and rigid enough, we assume it as a constant and attribute the variation of the interference phase δϕ (t) to the measurement arm L. Therefore, the vibration displacement ΔL(t) and velocity v(t) of the Raman mirror during the time period of [t, t + dt] is proportional to the difference of the phase shifts in interference signals by
According to Eqs. (
As shown in Fig.
It seems that we could solve the three unknowns of g, δϕRL1, and δϕRL2 from Eq. (
Though we cannot measure the absolute vibration of the mirror directly by MIRL, this method is still effective for the evaluation of the mirror vibration noise on AI. Denoting the standard deviation of the vibration noise induced phase instability on AIi, Raman mirrori, MI and gravity value as σΦAI, i, σΦvibM,i, σΦMI, and σΦg, respectively, we can derive the relationships between these Allan variances from Eq. (
Assuming the phase instability is independent of each other, we can solve
As to the independence of the phase instability, it could be verified by computing the tends of the Allan covariation in the following equation:[29]
We had introduced in the above section that the MIRL is an effective method for evaluating the contribution of vibration to the instability of the atomic gravimeter. Considering the MIRL is different from standard MI and has some unique characteristics, we will analyze its measurement accuracy quantitatively in the following section.
The Raman lasers used in an 87Rb atomic gravimeter are composed of two beams with wavelength of 780.24 nm and frequency difference of about 6.8 GHz, thus the interference signal of the MIRL has a beat signal of 6.8 GHz which is out of the photodetectors’ response range and can be neglected. Therefore, the MIRL can be approximately considered as an amplitude overlapping of two MI signals with tiny difference in laser frequencies. Furthermore, as shown in Fig.
In general, the intensity ratio η of the two Raman beams is typically around 1.5 – 2.5, which would lead to a periodic variation of interference signal envelope, as shown in Fig.
Generally, the Raman laser linewidth used in the atomic gravimeter is about 60 kHz and the relative frequency instability is about Δν / ν = 1.54× 10−10. Considering the relationship between phase and frequency is Δϕ = ∫ Δν dt, the phase measurement precision of the single-frequency laser interference measurement method is evaluated as δϕl,s = 2π Δν ΔL0 / c = 2.5 mrad. For the MIRL measurement method, the measurement error is dependent on the frequency instability of both lasers, thus can be evaluated as
For atomic gravimeter, the frequency difference of Raman lasers is scanned at a specific chirp rate of α ≈ 25.1 MHz / s in order to compensate the Doppler shift.[11] The chirped frequency difference not only affects the envelope position, but also leads to the phase change of the fine interference fringes, as shown in Fig.
The fringe movement due to frequency shift can be represented in phase as
Besides above characters of the MIRL, the air refractive index is also an important factor. Ignoring the variation of air composition, the air refractive index has a relative error of Δn/n = 10−8.[23] The measurement error induced by the change of the air refractive index is expressed as Δφair = 2π ΔnLair / (nλ0), where λ0 is the laser wavelength in vacuum, and Lair is the total arm length of MI in air. We evaluate the maximum measurement error caused by the change of air refractive index as 1.6 mrad from the maximum Lair ∼ 4.4 cm of the beat period length.
The total phase measurement error of the above factors is calculated by
We proposed an innovative method to measure the mirror vibration noise contribution of the atomic gravimeter based on a Michelson interferometer of Raman laser (MIRL). This MIRL is constructed with the atomic gravimeter intrinsic Raman beam combing a four-channel phase shift detector. Therefore, this vibration monitoring method has the advantages of simple, low cost, completely synchronized with the gravity measurement process, perfectly localized with the Raman beams and high measurement accuracy of 4 mrad or 10−10 g, etc. Considering this MIRL only can measure the relative motion between the beam splitter and bottom mirror, we propose a modified “AI–MI–AI” tri-interferometer system and “the triangle cap” method to measure the contribution of the mirror vibration on the instability of g measurement. Although this method cannot give the absolute vibration of the bottom mirror because the limit of the equivalent principle, its advantages of simple, low cost, effective and high-precision enable it an alternative method to evaluate the contribution of vibration noise to the instability of g measurement besides the commercial seismometer or accelerometer. Therefore, this paper is beneficial to the noise evaluation, miniaturization development, and field application of the atomic gravimeter.
[1] | |
[2] | |
[3] | |
[4] | |
[5] | |
[6] | |
[7] | |
[8] | |
[9] | |
[10] | |
[11] | |
[12] | |
[13] | |
[14] | |
[15] | |
[16] | |
[17] | |
[18] | |
[19] | |
[20] | |
[21] | |
[22] | |
[23] | |
[24] | |
[25] | |
[26] | |
[27] | |
[28] | |
[29] |