Quantum Sensors for Earth Observation

SR17_Weltraum_V8_Abstract_InfosysII.qxd 10.12.10 10:39 Seite 178

Quantum sensors for Earth observation

M. Gilowski and E.M. Rasel Institut für Quantenoptik, Leibniz Universität Hannover, Welfengarten 1, 30167 Hannover, Germany

Abstract mes [3-8]. This field represents an emerging We report on our development of various area of science quantum engineering, with a kinds of quantum sensors based on matter- high potential for a future technology and mul - wave interferometry for precise inertial sensing tidisciplinary applications. Thanks to an impres- of Earth`s rotation rate and gravity. By manipu- sive evolution and remarkable inventions, the lating cold or ultra cold atoms with coherent ultimate potential of matter-wave sensors is still light fields we realize several experimental plat- entirely open. For the closely related field of ato- forms for the detailed investigation of different mic clocks, the growth in performance was interferometer types and of their key technolo- exponential during the last decades! This is the gies. In an atomic gyroscope using cold Ru bi di - reason, why matter-wave sensors are conside- um atoms we demonstrate a sensitivity for red as one of the most promising fields to pro- rotations of 6x10-7 rad/s. In an atomic gravi- gress in metrology and fundamental tests. meter, which is under construction, we study Atomic gyroscopes and gravimeters provide a the source system based on ultra cold atoms in new tool for the precise detection of tiny for- a Bose-Einstein condensate. Finally, a miniatu- ces. The outstanding feature of these sensors rized and robust experiment using ultra cold is the precisely known scaling factor: there is atoms in a free fall environment is realized as a no need for calibration which predestines test-bed for matter-wave interferometry on these sensors for inertial references and for long timescales. These experiments pave the applications in the Système International. way in the direction of utilizing the technolo- The following report summarizes our activities gy of matter-wave interferometry for future and achievements in the field of matter-wave space applications. interferometry. We will describe the features of this technology by presenting three experiments which point out the development of Introduction atomic inertial sensors. The first one is an ato- Atomic quantum sensors are a key-technology mic gyroscope for the high-precision measure- for the ultra-precise monitoring of accelera- ment of rotations. The second one is a dual tions and rotations. These sensors evolved atomic gravimeter, which is under construction from a new kind of optics based on matter- and which provides a source of ultra cold waves rather than light-waves. Matter-wave atoms. The third one is a miniaturized trans- optics is still a young, but rapidly progressing portable experiment with ultra cold atoms in a science which recently generated sensational microgravity environment, which is performed Nobel-prize awarded inventions such as laser in the drop tower facility (ZARM) in Bremen. cooling and atom lasers [1,2]. These experiments are steppingstones in the The applications of atomic quantum sensors direction of the realization of future inertial are truly interdisciplinary, covering diverse and atomic sensors with sensitivities compared or important topics such as tests of fundamental even beyond current state-of-the-art devices. physics, the realization of SI-units, prospecting for resources, GALILEO technology, environment monitoring and major Earth-science the-

178 SR17_Weltraum_V8_Abstract_InfosysII.qxd 10.12.10 10:39 Seite 179

The gyroscope a subsequent 3D-MOT with a high flux of seve- The interferometric measurement of rotations ral 109 at/s. Using the moving molasses techni- is based on the Sagnac-effect [9], which indi- que 108 atoms with a forward drift velocity of cates that a phase shift is induced between 2.8 m/s and a temperature of 8 µK are laun- two interferometer paths which enclose an ched in each interferometry pulse. In a next area A, due to a rotation with the angular step the atoms are state- and velocity-selecti- velocity Ω. This phase shift is then given by vely transferred into the magnetically insensiti-

dfrot=4pEAΩ/hc, where E is the energy of the ve hyperfine state F = 1, mF = 0 (where F and

wave, h the Planck’s constant and c the speed mF are the quantum numbers for the total an - of light. Since this relation is also valid for light gular momentum and the Zeeman sublevel, res - as well as for matter-wave interferometers, the pectively) via a multi-stage preparation using high potential of gyroscopes based on atoms is precisely controllable laser manipulation. obvious. By comparing the phase shifts using In the interferometery section a symmetric Ram - the energies of matter and visible light, an im - sey-Bordé interferometer configuration is reali- provement in the order of 1011 for atom inter fe - zed by applying four so called p/2-light pulses rometers is in principle possible. Further more, [14]. This coherent beam splitting processes the sensor is also sensitive to accelerations [10]. are based on a Raman-transition between two In order to distinguish between phase shifts due hyperfine states which form the ground states to rotations and accelerations the sensor con- of the interferometer. With these light pulses sists of two interferometers allowing a differen- the matter-wave is split, redirected and finally tial measurement [10]. The basic schematic of recombined. In this way, the created interfero- the atomic gyroscope is sketched in Fig. 1. meter paths enclose an area of 8.6 mm2 leading to a high sensitivity for the Sagnac-effect. Two identical atomic sources [13] emit atoms In this interferometer configuration phase on flat parabolic trajectories into the interfero- shifts imprinted during the interferometric meter chamber, but with opposite launch di - cycle are translated into a change of the distri- rec tions. Each source consists of a two-dimen- bution in the two ground states [14]. Thus, we sional magneto-optical trap (2D-MOT) loading finally obtain the phase shift of each interfero-

Figure 1: Schematic overview of the gyroscope with its key-element sections. Laser beams for cooling and trapping as well as for the preparation, the interferometry and the detection process are represented by the arrows. More details can be found in reference [11,12].

179 SR17_Weltraum_V8_Abstract_InfosysII.qxd 10.12.10 10:39 Seite 180

Figure 2: The Interference fringes of the two interferometers are obtained by sweeping the phase of the beam splitting light field before the last pulse is applied.

Figure 3: Allan standard deviation of the combined signals of the two interferometers.

meter by using a state-selective fluorescence gain the contrast of the interferometer signal. detection, applied after the interferometry se - Furthermore, a reduction of the initial tempera- quence. We detect in the gyroscope approxima - ture of the ensemble would increase the effi- tely several 107 atoms. ciency of this process and thus motivates the Typical interference patterns of the two interfe- use of ultra cold atoms (see below). The diffe- rometers of the gyroscope are shown in Fig. 2. rence in contrast can be attributed to the align- We reach a contrast in the four pulse geome- ment of the spatial interaction position, which is try of 11% and 16%. The reduced contrast different for the four beam splitting pulses for can be attributed to a reduced efficiency of the the two interferometers, respectively. beam splitting process, which depends on the one hand on the temperature and the spatial Finally, we can infer the rotation sensitivity of width of the atoms and on the other hand on the gyroscope by measuring the phase noise of the spatial intensity profile of the beam split- the two interferometers. In Fig. 3 the Allan ting light field. A further selection of velocity standard deviation for the combined signal of classes in the preparation stage will in future the two interferometers is shown. Here, the

180 SR17_Weltraum_V8_Abstract_InfosysII.qxd 10.12.10 10:39 Seite 181

rotation signal can be inferred from the sum of on matter-wave interferometry is setup for the the two interferometer signals whereas the test of the Equivalence principle. For this purpo- acceleration might be calculated from the dif- se the free fall of two atomic species, Rubidium ferential signal. It is clearly visible that vibration and Potassium, will be compared. In this experi- noise coming from accelerations is highly sup- ment the aimed shot-noise limited sensitivity is pressed due to the dual measurement scheme. 10-9 g. This corresponds to the highest sensitivi- After averaging the combined signal over 300 s ty for the test of the Equivalence principle with a phase noise of 9.5 mrad is reached. This cor - laser cooled atoms so far [12]. responds to a reached sensitivity in the gyrosco- One of the essential aspects of the dual gravi- pe of 6x10-7 rad/s which is two orders of mag- meter is the source of ultra cold atoms to reach nitude below the Earth rotation rate. the demanded accuracy. The first step in this direction was reached by creating a Bose-Ein - Currently, this sensitivity is limited by the noise stein condensate (BEC) of 87Rb atoms in a near- of the detection process as well as inertial noise. infrared single beam optical dipole trap (ODT). The implementation of an optimized detection The ODT is formed from laser light with a wa ve - sys tem and an improvement of the vibration length close to 2 µm. Fur thermore, an additional isolation are under way and will further impro- constant confinement in the axial direction of ve the sensitivity by at least one order of mag- the trap is provided by a magnetic quadrupole nitude on the short-term scale. In combination field with a rather weak gradient of 10 G/cm. with novel techniques like the transfer of large The experimental description starts from a photon momentum leading to a larger transver- double MOT source system, which is related to sal atomic velocity and thus to larger enclosed the one explained in the above section. Here, areas [15], the aimed sensitivity might reach the 5 × 108 atoms are trapped in the 3D MOT. In 10-10 rad/s level for integration times of 1000s. the next step, 2×106 atoms are transferred into A further optimization is the reduction of the a single beam ODT at an initial temperature of atomic temperature, which besides the increa- 35 µK and a corresponding initial phase space sing of the beam splitter efficiency would density of about 1−2×10-5. By a following re - decrease systematic effects coming for instan- duc tion of the ODT laser intensity at constant ce from wave front distortions of the beam magnetic field gradient the evaporative coo- splitting light fields. The realization of such an ling is achieved. This allows for the creation of ultra cold atomic source for atom interferome- a nearly pure condensate with 104 atoms every try is described in the following section. 20 seconds as it is shown in Fig. 4. Furthermore, before starting the evaporation The ultra cold atom source sequence laser cooling in the ODT provides In a second experiment a dual gravimeter bas ed atomic samples with very low initial tempera-

Figure 4: Absorption images of the atomic cloud in the weak hybrid trap at different temperatures and their corresponding density profiles after a time-of-flight of 21 ms, showing the phase transition from a thermal gas in (a) to a bimodal density distribution in (b) to a quasi pure BEC with 104 atoms in (c).

181 SR17_Weltraum_V8_Abstract_InfosysII.qxd 10.12.10 10:39 Seite 182

tures of about 2 µK and initial phase space den - The BEC in microgravity – The QUANTUS sities as high as 10-2. These very favorable initi- project al values offer much potential for further im - The QUANTUS project (QUANTen Gase Unter prove ment in the production of a BEC with res - Schwerelosigkeit) started as a feasibility study pect to particle number and repetition rate. An of a compact, robust and mobile experiment adaption of the laser intensity ramp for forced for the creation of a BEC in a weightlessness evaporation to these strongly improved initial environment [16,12]. It was the drop tower in con ditions offers great potential for this me - Bremen, which was chosen as the microgravity thod to create BECs with particle numbers of environment, since it provides both good more than 105 atoms with a substantially im - accessibility and a better quality of microgravi- proved repetition rate. Moreover, this source ty than other platforms. This makes the expe- scheme demonstrates a matter-wave interfero- riment an ideal test bed for future space based meter source using ultra cold atoms. missions with ultra cold atoms. Besides the ultra cold atoms for improving the The capsule, which is released from the top of accuracy of a matter-wave interferometer long the evacuated tower, contains a complete BEC interrogation times are preferable, since the experiment as it is shown in Fig. 5. The realiza- induced inertial phase shifts in a matter-wave tion of this experiment required a massive mi ni - in terferometer scale quadratically with the in - aturization of the setup, which was made pos- ter rogation time for instance in a Mach-Zehn- sible by the development of atom chips [17]. der configuration [14]. If not only the atoms We combine this technology with a mirror but also the whole experiment itself would be MOT, which is loaded with roughly 1.3 × 10 7 in a free fall, the interrogation time could sig- atoms of 87Rb from the background gas. nificantly be extended. And this is one of the motivations for the following experiment. After 10 s of loading, the capsule is released and is free falling over a time of 4.7 s. Within this time a BEC of about 104 atoms is produ-

Figure 5: ZARM drop tower facility in Bremen (left). Capsule containing the BEC experiment (right). The future QUANTUS-II capsule will be launchable from a catapult. With this, the time of free fall can be almost doubled.

182 SR17_Weltraum_V8_Abstract_InfosysII.qxd 10.12.10 10:39 Seite 183

Figure 6: Interference pattern of a BEC in a two pulse interferometer on ground.

ced by the following procedure [16]: The ring the usability of BEC as a source for matter- atoms are further cooled in optical molasses wave interferometry, the coherence of the and transferred to an Ioffe-Pritchard trap on chip-based BEC is an important property to be the chip. By compressing the trap and radio studied. Furthermore, the precise preparation frequency–induced evaporation a condensate in the nK regime developed in the expansion

in the hyperfine state F = 2, mF = 2 is obtained. campaigns has to be proven sufficient for Finally, after switching off the current of the interferometry needs in increasingly long inter- atom chip the BEC is released and can be ob - ferometer sequences. Here, the reproducibility served by using the absorption imaging techni- of the interferometer over multiple drops is a que. During the long expansion over 1 s, the major challenge which will be met with a more atoms form a giant coherent matter wave that suitable choice of initial state. A coherent trans- is delocalized on a millimeter scale. fer of the BEC into the magnetically in sensitive

We have performed more than 190 experi- hyperfine state F = 2, mF = 0 would avoid the ments to demonstrate the feasibility of cohe- influence of parasitic magnetic effects. rent matter-wave experiments in microgravity, The implementation of this transfer is current- thus enabling the realization of a robust ato- ly under way. The design and potential of this mic source, which is required for the imple- technique will be studied in various experi- mentation of an atom interferometer. ments in 2011. The results and experience will First interferometer experiments with this devi- be extended to two species (Rubidium and Po - ce on ground have already been demonstrated tassium) in a new chip based drop tower expe- and will be performed in microgravity by the riment which is under construction and provi- end of this year. Here, Bragg scattering is used des improved atom flux and preparation. as the coherent beam-splitting and mirror process. On ground, interferometer sequences in the Mach-Zehnder and Ramsey geometry have Outlook been realized so far (see Fig. 6). The reported activities on atomic quantum sen - sors in the Institute of Quantum Optics show The former being the candidate for high sensi- the enormous potential of these devices for tivity measurements at long times while the the future application in Earth observation and latter is our candidate for coherence measure- fundamental physics. The presented atomic sen - ments. Since, our investigations aim at explo- sors serve as a platform for experiments aiming

183 SR17_Weltraum_V8_Abstract_InfosysII.qxd 10.12.10 10:39 Seite 184

at studying the relevant systematic errors and References limitations and in the next period to push these [1] W.D. Phillips, Rev. Mod. Phys., Nobel devices beyond current state-of-the-art. Fur ther- Lecture, 70, 3, (1998). more, tests of new quantum sensor concepts and strategies (continuous cold atom sensors, [2] W. Ketterle, Rev. Mod. Phys., Nobel Lecture, sensors based on degenerate ensembles, novel 74, 4, (2002). topologies) are on the way. Another important aim is a direct comparison [3] M. Cadoret, et al., Phys. Rev. Lett., 101, of different atomic inertial sensors, combined 230801, (2008). with the use of the best currently available geo - physical models for facilitating a precise evalu- [4] A. Bertoldi et al., Eur. Phys. J. D, 40, 271, (2006). ation of these instruments. The cooperation of experts on atom interferometer sensors with [5] P. Cheinet, et al., Dig. Conf. Precision ex perts from the field of Earth observation Electromagnetic Measurements (CPEM) 2004, will also permit the design and development London, U.K., 60-61, (2004). of improved inertial quantum sensors specifi- cally tailored for demanding applications in [6] D. S. Durfee, et al., Phys. Rev. Lett., 97, this field.The resulting advanced sensors will 240801, (2006). then offer many more exciting possibilities. In the long run the future devices will comple- [7] Dimopoulos, et al., Phys. Rev. Lett., 98, ment and can even replace the classical ones 111102, (2007). for monitoring mass changes in active regions (e.g. volcanoes) and also help to discriminate [8] G.M. Tino et al., Nuc. Phys. B, 166, 159-165, between man-made and natural effects (e.g. (2007). changing water reservoirs). The isolation of the sensor from Earths environ- [9] M.G. Sagnac, Compt. Rend. Des Sc. D. ment, will finally enhance the potential of mat- l’Acad. D. Sc., 157, 1410, (1913). ter-wave interferometry. In future satellite missions such instruments may test gravity fields as [10] T.L. Gustavson, et al., Class. Quantum Grav., gravimeters or gradiometers with high accuracy. 17, 2385, (2000). The in this report presented experiments are stu- dies of this novel and promising approach. [11] T. Müller, et al., Eur. Phys. J. D, 53, 273–281, (2009).

Acknowledgements [12] The homepage of the Institute of Quan - The project »Zukunftskonzepte für Schwere feld - tum Optics, http://www.iqo.uni-hannover.de, satellitenmissionen« is part of the R&D-Pro gram- (Au gust 2010). me GEOTECHNOLOGIEN. GEOTECHNOLOGIEN is funded by the German [13] T. Müller, et al., Phys. Rev. A, 76, 063611, (2007). Ministry of Education and Research (BMBF) and the German Research Foundation (DFG). [14] Ch.J. Bordé, Phys. Lett. A. 140, 10-12, (1989). We also thank the center for quantum engineering and space time research (QUEST). [15] H. Müller, et al., Phys. Rev. Lett., 100, Furthermore, we are grateful to the German 180405, (2008). Aerospace Center (Deutsches Zentrum fur Luft- und Raumfahrt e.V. – DLR). [16] Zoest, et al., Science 328 (5985), 1540- 1543, (2010).

[17] W. Hänsel, et al., Nature 413, 498, (2001).

184