A cold atom pyramidal gravimeter with a single laser beam

Abstract

We demonstrate a scheme for realizing a compact cold atom gravimeter. The use of a hollow pyramidal configuration allows to achieve all functions: trapping, interferometer and detection with a unique laser beam leading to a drastic reduction in complexity and volume. In particular, we demonstrate a relative sensitivity to acceleration of gravity (g) of 1.7 x 10-7 at one second, with a moderate laser power of 50 mW. This simple geometry combined to such a high sensitivity opens wide perspectives for practical applications (P. Bouyer and A. Landragin, patent number FR2009/000252, 2009).

Full text

Page 1

arXiv:1004.5290v1 [physics.atom-ph] 29 Apr 2010
A cold atom pyramidal gravimeter with a single laser beam
Q. Bodart,1S. Merlet,1N. Malossi,1F. Pereira Dos Santos,1P. Bouyer,2and A. Landragin1∗
1LNE-SYRTE, Observatoire de Paris, CNRS, UPMC,
61 avenue de l’Observatoire, 75014 Paris, France and
2Laboratoire Charles Fabry de l’Institut d’Optique, CNRS et Univ Paris Sud,
Campus Polytechnique, RD 128, 91127 Palaiseau Cedex, France
(Dated: 15 December 2009)
We demonstrate a scheme for realizing a compact cold atom gravimeter. The use of a hollow
pyramidal configuration allows to achieve all functions: trapping, interferometer and detection with
a unique laser beam leading to a drastic reduction in complexity and volume. In particular, we
demonstrate a relative sensitivity to acceleration of gravity (g) of 1.7 × 10−7at one second, with a
moderate laser power of 50 mW. This simple geometry combined to such a high sensitivity opens
wide perspectives for practical applications (P. Bouyer and A. Landragin, patent n◦FR2009/000252,
2009).

Page 2

2
Gravimeters, based on atoms interferometry, measure the Earth’s gravity as a phase shift between two paths of
matter waves [1]. Applications of such sensitive gravimeters cover numerous fields, from fundamental physics [2–5] to
navigation and geophysics. However transportable interferometers are required for foreseen applications in the field
of navigation and gravity field mapping [6]. In this paper, we show that an atom interferometer based on Raman
transitions [7] can be realized exploiting an hollow pyramid. This geometry enables a drastic reduction in complexity
of atomic gravimeters, replacing all laser beams (typically nine independent beams [1, 8]) by only one. Its permits
building much more compact instruments with moderate laser power and with performances comparable to state of
the art sensors. Pyramidal magneto-optical traps (MOT) [9] have been introduced to reduce the size and simplify
cold atoms experiments since they require a single laser beam. They have been used for different kinds of alkali atoms
in various experiments, with a hole at the vertex for generating continuous beams [10–12], for direct loading of 3D
MOTs [13], for quantum gases experiments [14] and eventually in microscopic pyramids on atom chips [15]. In our
experiment, the many reflections of the single incident beam onto the four mirrors of the hollow pyramid allow to
obtain the needed polarizations not only for trapping the atoms inside the pyramid but also for efficiently driving the
Raman transitions and for detecting the atoms [16].
FIG. 1. Experimental setup of the experiment; the two frequency collimated laser beam arrives from the bottom by one single
fiber and cools down 1.2 106atoms to 2.5 µK. At each measurement cycle, these two laser frequencies are detuned to realize
the Raman pulses of the interferometer. The total high from the top of the pyramid to the detection is of 15 cm.
We first briefly describe the experimental setup, pictured in Fig. 1, and the principle of operation of the gravimeter.
We load a magneto optical trap directly from a vapor of 3 × 10−9hPa of Rubidium 87 (Rb). At each cycle, the
pyramidal MOT traps typically 4 × 106atoms in 360 ms. After a molasses stage of 20 ms, we switch off the light
adiabatically and let the atoms fall. We have measured by Raman velocimetry [17] a temperature of 2.5 µK, as
low as the one obtained with the usual configuration based on six independent laser beams. In order to reduce the
sensitivity to magnetic field, we select atoms in the |F = 1,mF = 0? state, using a sequence of a micro-wave and
pusher beam pulse at the beginning of the free fall. As soon as the atoms have left the pyramid, we perform a velocity
selection in the vertical direction, leaving about 3×105cold atoms within state |F = 1,mF= 0?. Then we realize an
interferometer with a π/2 − π − π/2 sequence [7], with an interrogation time of up to 80 ms. Thank to the internal
state labeling technique [18], the interferometer phase shift is extracted from the measurement of the population in
both output ports of the interferometer by laser-induced fluorescence on their associated internal states |5S1/2,F = 1?
and |5S1/2,F = 2?. The total sequence duration lasts 560 ms. In addition, the entire experiment is placed onto a
passive isolation platform in order to reduce the influence of spurious vibrations.
The use of a single laser beam leads to many simplifications in the laser system as it avoids all optical elements
for the splitting, transport and independent power control of many beams. This single beam is composed of two
frequencies, whose difference corresponds to the microwave transition of the Rb ground levels (i.e. 6.8 GHz). The
laser system consists in a simplified version of the one already described in detail in [19]. We briefly recall here
the main features. Two extended-cavity diode lasers are tuned respectively close to the |5S1/2,F = 1? → |5P3/2?
and |5S1/2,F = 2? → |5P3/2? transitions at 780 nm. The laser frequencies are independently controlled thanks to

Page 3

3
a versatile electronic system, based on beatnote measurements and frequency to voltage conversion. This system
allows fast changes of the laser frequencies, from very close to resonance for the cooling phase to detuned by 151 MHz
with respect to |5P3/2,F
beamsplitter cube, and finally injected into the same polarization-maintaining fiber with a 1:4 ratio. At the output
of the fiber, the beam has a total power of 50 mW and is circularly polarized. It is collimated with a waist of 14 mm
(1/e2radius) before entering the vacuum chamber from the bottom.
The inverted pyramid was manufactured [20] out of two glass corner cubes and two glass isosceles rectangular
prisms glued together, on a pedestal. The corner cubes are glued on the four ridges in order to obtain right angles
between the opposite faces inside the pyramid. We have controlled the angle to be 90◦within one arc minute. The
pyramid base is a square of 20 × 20 mm2area. The pyramid is attached to the upper flange pointing upward. Inner
faces are dielectrically coated for maximum reflection at 45◦and for equal phase-shift between the two orthogonal
polarizations. Therefore, reflections of the single beam onto the four inner faces of the pyramid creates the required
3D σ+/σ−polarization configuration for the trap inside the pyramid [9]. In addition, outside the pyramid, the laser
field configuration allows driving either σ+/σ+or σ−/σ−vertical velocity selecting Raman transitions [7] needed to
realize the interferometer and the detection.
′= 1? for the interferometer. These two laser beams are superimposed onto a polarizing
?
??
?
?
?
FIG. 2. Interferometer fringe pattern for a total interferometer time of 80 ms. Each point corresponds to a single measurement.
We now characterize the performances of the interferometer using the pyramid. The measurement of g is extracted
from the frequency chirp of the Raman laser difference that compensates exactly the time-dependent Doppler shift,
as described in [19]. Figure 2 displays the fringe pattern obtained for the maximal total interferometer duration of
80 ms and a Raman pulses sequence τ − 2τ − τ, with τ = 9 µs. The fringe pattern is obtained by scanning the
transition probability between hyperfine quantum states versus the interferometer phase-shift induced by a phase
jump on the Raman lasers phase difference between the second and the third pulse. The vibration noise is reduced
by using corrections obtained from the signal of a highly sensitive seismometer [8]. We obtain a contrast of 19 % and
a signal to noise ratio of 23 shot to shot at half fringe.
Continuous gravity measurements have been recorded during 50 hours and agree with the results of predicted
Earth’s tides [21], as displayed on Figure 3. In order to track fluctuations of gravity, we servo-lock the value of the
frequency chirp to compensate for the Doppler effect at any time. The error signal of the servo-loop is calculated
from the difference of alternative measurements on the two sides of the central fringe.
Figure 4 shows the Allan standard deviation of the relative g measurements when subtracting the tides signal
calculated from the model. The sensitivity improves as 1.7×10−6m.s−2.τ−1/2up to 1000 s, where τ is the measurement
time. This correspond to a relative short term sensitivity to g at one second of 1.7 × 10−7). This performance lies
about one order of magnitude above best atomic gravimeters [5, 8]. The technical limits were attributed to residual
vibrations on the platform on one hand and spurious phase noises on the micro-wave reference frequency used for
the Raman transitions on the other hand. Long term stability reaches a flicker floor below 5 × 10−9g. Its limit is
attributed to fluctuations of the systematic error corresponding to Raman laser wave-front distortions.
The wavefront distortion from the reflections leads to a systematic effect on the interferometer phase. We evaluate
this phase shift as a function of the transverse expansion of the cloud by changing the temperature during the molasses
phase. We found a linear dependance of about 5×10−7m.s−2per µK. Finally, we have also estimated the modification
of the effective wave-vector of the Raman transition by the Rb background pressure compared to perfect vacuum,

Page 4

4
?
?
?
?
FIG. 3. Variation of Earth’s gravity as a function of the time. The dots present experimental data averaged over 22 minutes.
The line corresponds to predicted Earth’s tides.
?
?
?
?
?
FIG. 4. Allan Standard deviation of relative gravity measurements corrected from the Earth’s tides.
which leads to a systematic error of the level of 6 × 10−8m.s−2.
These performances, which have been obtained despite a moderate contrast, can be improved in many ways.
Higher atom number, which enables a thinner velocity class selection by the Raman transition, can be achieved with
higher laser power. This will also allow for a larger laser beam size and Raman detuning, thus reducing transverse
laser intensity variations and the decoherence by spontaneous emission (3.6% in this experiment). Improvement of
the pyramid reflector will reduce the intensity and phase inhomogeneities across the beam. Because of the radial
expansion of the cloud, the transverse motion of the atoms or a change in their initial position, these homogeneities
result in a reduction of the contrast, from 31% for 2T = 2 ms to 19% for 2T = 80 ms. In our experiment, the better
contrast was obtained when atoms were located far from the dark lines corresponding to the edges of the pyramid
faces.
In conclusion, we demonstrated that a pyramidal reflector allows to realize a compact atomic gravimeter with only
one laser beam of moderate power. The short term sensitivity was limited by technical noises, which can be reduced
to 6 × 10−8g.τ−1/2, as demonstrated in another experiment with similar parameters (interrogation time, vibration
level of the ground, repetition rate, number of atoms) [22]. The fundamental limit, which is due to the atomic shot
noise, has been estimated to be 3 × 10−8g.τ−1/2, and can be pushed down with larger laser power. Compared to
commercial ballistic corner cube gravimeters, these performances are already comparable with compact models [23],
and can reach performances at the level of the state of the art [24].
The reduction in the complexity leads to a drastic reduction of the volume of the physical package from few hundred

Page 5

5
to few liters. As it also simplifies the optical bench and requires only a moderate optical power, similar reduction of
the size of the optical part is also expected compared to these of standard cold atom experiments (for example by
using an all fibered system [25]). Moreover, this gravimeter is scalable: increasing or decreasing the size of the pyramid
and the height of free fall allows to change in the same time the expected performances by changing the number of
collected atoms and the interrogation time. Such a system, combined with a compensation of the residual vibrations
by correlation with the signal of a mechanical seismometer [26], opens the way for practical field applications in
gravimetry and more generally in inertial force measurements with atom interferometry.
We would like to thank the Institut Francilien pour la Recherche sur les Atomes Froids (IFRAF) and the European
Union (EuroQUASAR/IQS project) for financial supports. Q. B. thanks the CNES for supporting his work.
∗arnaud.landragin@obspm.fr
[1] A. Peters, K. Y. Chung, S. Chu, Nature, 400, 849-852 (1999).
[2] G Genev` es, P. Gournay, A. Gosset, M. Lecollinet, F. Villar, P. Pinot, P. Juncar, A. Clairon, A. Landragin, D. Holleville,
F. Pereira Dos Santos, J. David, M. Besbes, F. Alves, L. Chassagne, and S. Top¸ cu, IEEE Trans. on Instr. and Meas. 54,
850-3 (2005).
[3] J. B. Fixler, G. T. Foster , J. M. McGuirk, M. A. Kasevich, Science Magazine 315, 74 (2007).
[4] G. Lamporesi, A. Bertoldi, L. Cacciapuoti, M. Prevedelli, and G. M. Tino, Phys. Rev. Lett. 100, 050801 (2008).
[5] H. M¨ uller, S. Chiow, S. Herrmann, S. Chu, K. Y. Chung, Phys. Rev. Lett. 100, 180405 (2008).
[6] R. E. Bell, V. A. Childers, R. A. Arko, D. D. Blankenship, and J. M. Brozena, Journal of Geophysical Research 104,
15281-15292 (1999).
[7] M. Kasevich, and S. Chu, Phys. Rev. Lett. 67 181 (1991).
[8] J. Le Gou¨ et, T.E. Mehls¨ aubler, J. Kim, S. Merlet, A. Clairon, A. Landragin, and F. Pereira Dos Santos, Appl. Phys. B
92, 133144 (2008).
[9] K. I. Lee, J. A. Kim, H. R. Noh, and W. Jhe, Optics Let. 15, 1177-1779 (1996).
[10] J. M. Kohel, J. Ramirez-Serrano, R. J. Thompson, L. Maleki, J. L. Bliss, K. G. Libbrecht, J. Opt. Soc. Am. B 20, 1161-1168
(2003).
[11] A. Camposeo, A. Piombini, F. Cervelli, F. Tantussi, F. Fuso, E. Arimondo, Optics Communications 200, 231-239 (2001).
[12] J. Arlt, O. Marag` o, S. Webster, S. Hopkins and C. J. Foot, Optics Communications 157, 303-309 (1998).
[13] R. S. Williamson III, P. A. Voytas, R. T. Newell, and T. Walker, Optics Express 3, 111-117 (1998).
[14] M. Vangeleyn, P. F. Griffin, E. Riis, and A. S. Arnold, Optics Express 17, 13601-13608 (2009).
[15] S. Pollock, J. P. Cotter, A. Latiotis, E. A. Hinds, Optics Express 17, 14109-14114 (2009).
[16] P. Bouyer and A. Landragin, patent n◦FR2009/000252 (2009).
[17] M. Kasevich, D.S. Weiss, E. Riis, K. Moler, S. Kasapi and S. Chu ”Atomic velocity selection using stimulated Raman
transitions”, Phys. Rev. Lett. 66 2297 (1991).
[18] Ch. Bord´ e, Phys. Lett. A 140 10-12 (1989).
[19] P. Cheinet, F. Pereira Dos Santos, T. Petelski, J. Le Gou¨ et, J. Kim, K.T. Therkildsen, A. Clairon and A. Landragin, Appl.
Phys. B 84, 643-646 (2006).
[20] The pyramidal reflector have been realized by Optique de Pr´ ecision J.Fichou, 30 rue de la Garenne, 94260 Fresnes, France.
[21] The model was carried out by applying the PETGTAB software, version 3.01, produced by G. Wenzel, Institut f¨ ur
Erdmessung, University of Hannover, Germany.
[22] A. Gauguet, B. Canuel, T. L´ ev` eque, W. Chaibi, A. Landragin, Phys. Rev. A 80, 063604 (2009).
[23] J.M. Brown, T.M. Niebauer, B. Richter, F.J. Klopping, J.G. Valentine, and W.K. Buxton, Eos Trans. AGU 80, 355 (1999).
[24] T. M. Niebauer, G. S. Sasagawa, J. E. Faller, R. Hilt and F. Klopping, Metrologia 32, 159-180 (1995).
[25] R.A. Nyman, G. Varoquaux, F. Lienhart, D. Chambon, S. Boussen, J.-F. Cl´ ement, T. M¨ uller, G. Santarelli, F. Pereira
Dos Santos, A. Clairon, A. Bresson, A. Landragin, P. Bouyer, Appl. Phys. B 84, 673 (2006).
[26] S. Merlet, J. Le Gou¨ et, Q. Bodart, A. Landragin et F. Pereira Dos Santos, P. Rouchon, Metrologia 46, 87-94 (2009).