Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

NASA’s Cold Atom Lab (CAL): system development and ground test status

NASA’s Cold Atom Lab (CAL): system development and ground test status www.nature.com/npjmgrav ARTICLE OPEN NASA’s Cold Atom Lab (CAL): system development and ground test status 1 1,2 1 1 1 Ethan R. Elliott , Markus C. Krutzik , Jason R. Williams , Robert J. Thompson and David C. Aveline We report the status of the Cold Atom Lab (CAL) instrument to be operated aboard the International Space Station (ISS). Utilizing a 87 39 41 compact atom chip-based system to create ultracold mixtures and degenerate samples of Rb, K, and K, CAL is a multi-user facility developed by NASA’s Jet Propulsion Laboratory to provide the first persistent quantum gas platform in the microgravity conditions of space. Within this unique environment, atom traps can be decompressed to arbitrarily weak confining potentials, producing a new regime of picokelvin temperatures and ultra-low densities. Further, the complete removal of these confining potential allows the free fall evolution of ultracold clouds to be observed on unprecedented timescales compared to earthbound instruments. This unique facility will enable novel ultracold atom research to be remotely performed by an international group of principle investigators with broad applications in fundamental physics and inertial sensing. Here, we describe the development and validation of critical CAL technologies, including demonstration of the first on-chip Bose–Einstein condensation (BEC) of Rb with 39 87 41 87 microwave-based evaporation and the generation of ultracold dual-species quantum gas mixtures of K/ Rb and K/ Rb in an atom chip trap via sympathetic cooling. npj Microgravity (2018) 4:16 ; doi:10.1038/s41526-018-0049-9 INTRODUCTION The persistent free fall condition of low Earth orbit is recognized as the next natural destination for cold atom work to take The thermal, random motion of atomic gases fundamentally limits advantage of microgravity. Proposals include both satellite-based their free space measurement time and restricts their manipula- 21,25–28 missions and experiments to operate aboard the Interna- tion. Forty years of breakthroughs in the creation of ultracold 29–34 tional Space Station (ISS). The ISS becomes particularly quantum gases have resulted in standardized cooling techniques attractive when compared to a satellite mission both in terms of and technologies that are mature to the point of supporting large cost and the constant human presence, which enables instrument 1–3 4–6 scale dedicated facilities, portable field-ready devices, and modifications and upgrades. Leveraging these advantages, the 7–9 commercially offered components. In recent decades, national Cold Atom Lab (CAL) is designed to provide the first ultracold consortia and experimental groups seeking still colder tempera- quantum gas experiment aboard the ISS by utilizing an apparatus tures and extended measurement times have successfully pushed developed, assembled, and qualified by NASA’s Jet Propulsion ultracold atoms and related technologies into the free fall Laboratory (JPL). CAL is intended to operate as multi-user facility 10–12 13,14 conditions of drop towers, zero-g airplanes, and sub- for principal investigators to remotely study ultracold and 15–18 87 39 41 orbital launch vehicles. Here, two substantial benefits are quantum degenerate samples of Rb, K, and K, including realized. The first is access to a new parameter regime of quantum dual-species mixtures of Rb and K. In this article, we report the gases at picokelvin temperatures and ultra-low densities. Broadly, system development and test status of the CAL instrument, highlighting the successful production of dual species ultracold this is achieved by removing the asymmetric tilt that gravity gases in CAL’s ground testbed facility. introduces along one direction of the confining potential used to Conceptually, the CAL system consists of three primary trap and cool atoms. It is then possible, for example, to subsystems: the science module, electronics, and the laser and decompress these traps far beyond what is achievable on the optical distribution system. In order to support reliability, repair, ground. Secondly, the wave nature of ultracold atoms released integration, and replacement of components, all three subsystems from these traps is the basis for inertial sensing atom inter- leverage commercially offered components in an architecture ferometers, the precision of which scales as the square of their free intended to be compact, simple, and modular. This modular space evolution time. Recent milestones of these ruggedized, design has not only allowed parallel development of different autonomous devices operating in reduced gravity include the subsystems but also various versions of the same subsystem. In generation of Bose–Einstein condensates (BEC) and the demon- particular, CAL’s ground testbed currently supports multiple stration of matter-wave interferometry. These pathfinders physics packages, which are based on a commercially available underline both the technical feasibility and the future potential dual-cell vacuum chamber featuring an atom chip. CAL’s ground for high-precision fundamental physics applications including testbed has also developed two distinct “science modules,” the searches for dark energy, quantum tests of Einstein’s equiva- project designation for physics packages prepared specifically in 21,22 23,24 lence principle, and long-baseline gravity wave detectors. the flight configuration of robustly integrated optics, magnetic 1 2 Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91109, USA and Humboldt-Universität zu Berlin, Newtonstr. 15, 12489 Berlin, Germany Correspondence: David C. Aveline (david.c.aveline@jpl.nasa.gov) Received: 18 January 2018 Revised: 16 April 2018 Accepted: 30 May 2018 Published in cooperation with the Biodesign Institute at Arizona State University, with the support of NASA NASA’s Cold Atom Lab (CAL): system development and ground ER Elliott et al. procedure loads a limited number of K atoms into the chip trap and cannot afford the atom losses inherent to removing the hottest potassium atoms through a direct evaporative process. Instead, we evaporate only Rb, which in turn sympathetically cools trapped potassium. This cooling is accomplished specifically with a microwave evaporation scheme, transferring the most energetic |22⟩ Rb to the anti-trapped |11⟩ state, a state transition 37–45 that is separated by more than 6.8 GHz. Because the spacing 41 39 of the similar hyperfine states in K and K are 254 MHz and 462 MHz, respectively, microwave signals at 6.8 GHz address only Rb. This is in contrast to a more typical RF evaporation scheme, which instead operates on states with frequency spacings that are almost exactly the same in Rb and either species of bosonic potassium, specifically transitioning trapped |22⟩ atoms to the more weakly trapped |21⟩ state before arriving at the non- magnetic |20⟩ state or anti-trapped |2−1⟩ state. However, it follows that microwave evaporation yields the comparative disadvantage of not immediately removing any |21⟩ Rb atoms, should they also exist in the trap. The source of the |21⟩ impurity during microwave evaporation of |22⟩ Rb has been suggested to 37,39 result from either inelastic spin-changing collisions or an ejected |11⟩ atom moving to a point of higher magnetic field where the microwave knife becomes resonant with the |11⟩ to | 38,40,44 21⟩ transition. Consistent with either of these possible cases is that the |21⟩ impurity is observed to emerge during the evaporation process, and not simply as the result of inefficient optical pumping. If not addressed, these |21⟩ atoms can act as an additional heat load to be sympathetically cooled or potentially remove potassium atoms via energy released from interspecies collisions that do not conserve electronic spin states. Our microwave evaporation protocol incorporates five linear fre- quency ramps that at low temperatures are interspersed with brief clearing stages at a second fixed frequency (6.846 GHz) 87 39 Fig. 1 False color absorption images of Rb and K after 3 ms of resonant with the Rb |21⟩ to |11⟩ transition at the 8.4 G trap expansion from an atom chip trap at five different stages of bottom to recurrently remove the unwanted |21⟩ atoms. Begin- microwave-induced evaporation of rubidium and the corresponding ning 60 MHz above the bare hyperfine transition at 6.834 GHz, sympathetic cooling of potassium. Higher temperature/atom these five ramps together last a total of 1.5 s, with up to one number data are shown for illustrative purposes. The K atoms hundred equally spaced jumps to 6.846 GHz. loaded onto the chip are initially so diffuse that our imaging fails to Following evaporation, we decompress the trap via a 70% detect them. The last stage shows 70,000 K atoms at a reduction of the the bias fields in 100 ms before release and temperature of 1.6 μK and phase space density (PSD) of 0.05. Rb 39 87 numbers in the final stage are 155,000 atoms at a temperature of imaging. Initial results of sympathetically cooling K with Rb 39 4 1.6 μK and a PSD of 0.025 yield K temperatures of 170 nK with 1.1 × 10 atoms and a phase space density (PSD) of 0.6. For illustrative purposes, Fig. 1 shows a coils, two CMOS cameras, microwave and radio frequency (RF) higher temperature/atom number example through five stages of emitters, a water-based cooling loop, and magnetic shields. evaporation. Additionally, our system has cooled K to degen- Details of the science module and physics package can be found eracy, with current results yielding an 11% condensate fraction in the “Materials and Methods” section of this paper. This modular out of 4200 total atoms at 76 nK and a PSD of 3.5. The larger 41 87 39 87 architecture also allows more readily available prototype/ground collisional cross section of K/ Rb compared to K/ Rb support systems of optics and electronics to test the functionality increases not only the comparative evaporation efficiency, but of the science modules during development of the flight models. also yields more effective sympathetic cooling during the initial Specifically, all degenerate and ultracold atom results reported in loading of the magnetic quadrupole trap. Repeating the above this article use the flight science module, flight equivalent lasers/ procedure without loading potassium produces Rb BECs with optical fibers, and ground support electronics as the flight 3×10 total atoms and a condensate fraction of 10%, comparable electronics were finalized. The electronics described in Materials to the performance of our Rb BECs formed using standard RF and Methods reflect these ground versions. evaporation (5 × 10 atoms with 10% BEC fraction). To the best of our knowledge, these results mark the first use of an atom chip to evaporate Rb to condensation using microwaves, sympatheti- RESULTS cally cool either bosonic potassium isotope, and produce We have successfully produced degenerate samples of Rb and degenerate K. 41 39 K in addition to nearly degenerate K in the atom chip While the first group to demonstrate interspecies sympathetic magnetic trap within CAL’s flight science module. Beginning with cooling to a BEC did not include techniques to directly remove | 87 39 87 43 the loading of the atom chip trap from dual species Rb/ Kor 21⟩ Rb atoms, their subsequent implementation of a counter- 87 41 Rb/ K in 3D and 2D magneto-optical traps (MOTs), the measure against the |21⟩ state increased evaporation efficiency to 45,46 production of a Rb BEC follows using either RF or microwave- allow the first mixture of dual species BECs. All later induced evaporation though the selective rejection of the most implementations of Rb microwave evaporation have incorpo- energetic Rb atoms via transfer from the low magnetic field- rated an additional frequency targeting the |21⟩ atoms through- 38– seeking |F = 2, m = 2⟩ state to a high-field-seeking, anti-trapped out the entire evaporation ramp, either continuously operating 45 37 state. As described in the section “Materials and Methods,” our or repeatedly pulsed. As noted, CAL employs a variation of the npj Microgravity (2018) 16 Published in cooperation with the Biodesign Institute at Arizona State University, with the support of NASA 1234567890():,; NASA’s Cold Atom Lab (CAL): system development and ground ER Elliott et al. Fig. 2 Necessity of two frequencies to remove the |21⟩ impurity during microwave evaporation of |22⟩ Rb to BEC in a harmonic magnetic trap with trap bottom B . If a |21⟩ population exists in thermal equilibrium with |22⟩ atoms, it will form a cloud with a larger mean-square cloud size than the |22⟩ cloud due to weaker coupling to the magnetic field. (a) Maximum total energy of a trapped atom in either state exceeds μ B , where μ is the Bohr magneton, and a single frequency (green arrow) that evaporates the most energetic |22⟩ atoms to the anti- B 0 B trapped |11⟩ state will also remove the hottest |21⟩ atoms. (b) The maximum total energy is less than μ B , and the single |22⟩ evaporation B 0 frequency is too high to affect the |21⟩ atoms. (c) A second frequency (yellow arrow) resonant with the |21⟩ to |11⟩ transition at B removes all trapped 21 atoms latter technique. However, at higher temperatures, we find it is resonances to control differential center-of-mass distributions of 49,50 possible to evaporate sufficiently with only a single frequency. We dual-species quantum gas mixtures, bubble-shell geometries note that this is due to a combination of two effects that exist for | for Bose–Einstein condensates, as well as few-body systems in 21⟩ and |22⟩ atoms simultaneously confined in a harmonic new temperature and density regimes that are prerequisites for magnetic trap with trap bottom B . First, a |21⟩ population is more the next generation of Efimov experiments. After the conclusion weakly coupled to the magnetic field, experiences a looser of these experiments, the instrument will either enter a confinement, and when thermal equilibrium is assumed, forms a decommissioning phase or, through the use of orbital replace- larger cloud compared to a |22⟩ population. Secondly, a |21⟩ atom ment units, receive novel, upgraded capabilities for extended at the same magnetic field as a |22⟩ atom requires a lower mission operation and new microgravity experiments of advan- frequency microwave field to evaporate to the |11⟩ state. cing scope and complexity. Quantitatively, it is easily shown for this case that if a |22⟩ atom with total energy E relative to its trap bottom can climb the trap MATERIALS AND METHODS wall to a magnetic field of B, related by E = m g μ (B − B ) where F L B 0 μ is the Bohr magneton and g is the Landé g-factor, then a |21⟩ Science module B L atom with the same total energy E can reach a higher magnetic The physics package resides inside CAL’s science module, which also field given by 2B − B . If the most energetic |22⟩ atoms reach B,a houses all the hardware that maintains the appropriate local environment of magnetic fields, electromagnetic fields (optical, RF and microwave), as microwave field with frequency ω + 3Bμ /(2ℏ), where ω /2π is hf B hf well as thermal and mechanical stability (Fig. 3). Forming the outer the zero field Rb hyperfine splitting, is required to transition boundary of the science module is a dual layer magnetic shield, which them to the |11⟩ state. Comparison to the frequency that mitigates the significant and regularly modulated external magnetic fields evaporates the most energetic |21⟩ atoms, given by ω + (2B − hf experienced on the ISS. A bulkhead on the side of shield holds all B )μ /ℏ, yields a critical microwave frequency, ω = ω + 3B μ /ℏ, 0 B c hf 0 B necessary optical, electrical and thermal feedthroughs for the physics above which a single frequency is sufficient to evaporate |22⟩ package, including the water cooling loop, seven optical fibers (780 nm atoms while still decreasing the |21⟩ population, but below which cooling, 780 nm repump, 767 nm cooling and repump, imaging, optical a second frequency becomes necessary to remove the remaining | pumping, 2D-MOT push, and Bragg beam), power for the magnetic coils, 21⟩ impurity (Fig. 2). Similarly, there is also a critical total energy of atom chip currents, alkali metal dispensers, ion pump, two cameras (plus a trapped atom in either state relative to its respective trap datalinks), and RF/microwave signals. CAL’s physics package (Fig. 3a) is based on ColdQuanta’s commercially bottom given by μ B . Following this simplified model, a harmonic B 0 7,36 offered RuBECi vacuum chamber, modified for flight and CAL-specific magnetic trap with a higher trap bottom would therefore have a science objectives. These modifications include a second atomic source for relatively larger region where two frequencies are necessary for potassium, specialized anti-reflective coatings, a unique configuration of efficient microwave evaporation. atom chip conductive paths, as well as features for improved electrical, thermal, and mechanical integrity. The physics package consists of a dual glass-cell vacuum chamber, with one cell (source cell) housing two 12 mm DISCUSSION alkali metal dispensers (AMD, SAES) and a non-evaporable getter (NEG, Following the work presented here, one science module will SAES ST175), while the second (science cell) holds an ultra-high vacuum remain at JPL facilities for PI-specific experiment development and with an atom chip forming its ceiling. In order to maintain pressures below −10 10 Torr in the science cell, a miniaturized 2 L/s ion pump and a second further optimization of dual species quantum gas production, NEG are implemented in a six-flange stainless steel vacuum intersection. while the primary science module is integrated into the fight The source cell is surrounded by four rods lined with permanent magnets instrument for acceptance testing and final ground verification as for the production of overlapped dual species 2D-MOTs. A smooth silicon an autonomous instrument before delivery to the launch site. chip with a 0.75 mm diameter pinhole provides differential pumping with Following CAL’s arrival at the ISS, there will be an initial installation the science cell. This pinhole also allows the propagation of cold atoms phase by the onboard crew, including its first power up and basic from the 2D-MOT in the source cell to the 3D-MOT in the science cell. A communication tests between the instrument and the Ground push beam is directed along the axis of this pinhole from below the source Data System (GDS). In the subsequent commissioning phase, the cell, while partial reflection of this beam’s edges from the silicon surface 36,53 flight system performance will be tested and optimized for additionally boosts the loading rate of the 3D-MOT. The CAL science cell is surrounded by ten rectangular-shaped magnetic microgravity with the option of data downlinked to the GDS in coils, housed in an anodized aluminum structure that couples to the water real time. In the nominal operation phase, the CAL instrument will cooling loop via eight thermal straps. These enclosed coils produce the be run with a duty cycle of 1 min for up to 8 h per day during crew fields necessary for the MOT, magnetic transport, transfer into a chip trap, sleep, to minimize vibrations. These operations will be conducted and tuning near Feshbach resonances. The coils are capable of generating through a collaboration of the CAL PIs and the CAL team based at magnetic bias fields along each Cartesian axis. Along the x-axis, JPL, with each PI allocated schedule for their experiments to be independent coil control allows anti-Helmholtz configurations for mag- run. The proposed research will investigate applications of atom netic trapping gradients. Two separate pairs of vertically offset x-coils interferometry for future precision measurements in the areas of translate the zero-field position from the MOT location to the atom chip. 47,48 Earth observation and fundamental physics, Feshbach Incorporated into these upper x-coils (the transfer coils) are two turns of Published in cooperation with the Biodesign Institute at Arizona State University, with the support of NASA npj Microgravity (2018) 16 NASA’s Cold Atom Lab (CAL): system development and ground ER Elliott et al. Fig. 3 Layered overview of the CAL science module. (a) The physics package and magnetic coil assembly. MOT coils are shown in red, with the central axis of coils oriented along the x-axis. Transfer coils (in green) are vertically offset from the MOT coils. (b) Aluminum enclosure for the physics package securing the cameras and optical fiber collimators. Collimators and mirrors are emphasized in order to illustrate beam geometry, with red/orange arrows indicating effective directions of beam propagation. Two sets of 3D-MOT beams propagate diagonally in the y–z plane, while the third propagates along the x-axis through the MOT coils. Input for the optical pumping beam path (orange arrow) is positioned above the x-axis MOT beam collimator. The collimators opposite the two cameras provide light for absorption imaging, while the collimator below the through-chip camera also provides the push beam. The remaining two collimators surrounding the source cell send light to the 2D-MOT. Opposite the 2D- and 3D-MOT beam collimators are retro-reflecting mirrors (the retro-reflecting mirror for the x-axis MOT beam is not visible from this orientation). (c) The outside of this aluminum structure is bracketed to the water cooling loop and further enclosed by the magnetic shields with feed through connections from the other subsystems of the CAL instrument. Fully assembled, the dimensions are 46 cm × 30.5 cm × 58.5 cm, at a mass of 45 kg wires designated as the fast Feshbach coils (FFC). In conjunction with the the source cell directs a vertical beam through the center of the physics package to serve as either the push beam or the absorption imaging beam transfer coils in Helmholtz configuration (producing up to 325 G), the FFC for the through-chip camera. In order to provide efficient optical pumping, ramps the x-bias field by 3G within 50 μs. The upper side of the coil an additional dedicated output coupler and quarter wave plate on the side housing attaches to a breakout board for electrical connections to the of the science cell produces a slightly diverging beam that is roughly 1 cm atom chip, a microwave antenna, and a RF antenna. This breakout board in diameter at the location of the atoms. also concentrically secures these loop-style antennas a few mm above the atom chip (Fig. 5). Within the science cell, we have obtained a 10 s lifetime in the tight trap Laser and optical distribution system used for forced evaporation, where (ω , ω , ω )= 2π × (300, 1450, 1490) and x y z The optical distribution system of CAL, on the ground and in flight, is a variable hold time is introduced following three evaporation stages. A based on commercially available lasers and components to create an all variable hold time in the same trap following BEC production yields a optical fiber-based distribution system (Fig. 4). We operate one laser condensate lifetime of 2 s. Limitations to the trap lifetime result from noise system for both bosonic potassium isotopes at 766.701 nm, and an on the current drivers sourcing the currents for the magnetic trap (both the additional system for Rb at 780.24 nm. Laser cooling light is sourced by chip traces and coils) and vacuum quality. The flight electronics exceed the external cavity diode lasers (New Focus Vortex Plus) with a ruggedized performance of ground-use commercial chip current drivers, as the ground design for flight. Each laser features an internal 30 dB optical isolator and chip drivers achieve a lifetime of 8 s in the same configuration. We find that pigtailed optical fiber output coupler. The output light is guided through the lifetime is strongly influenced by the operation of the dispensers within single-mode polarization maintaining fibers (Corning PM 850) and fiber the source cell, although any related change in vacuum quality is below the splitter arrays from Evanescent Optics. range of the ion pump gauge. Continuous operation of both dispensers will For each atomic element, we use three separate lasers and a tapered gradually degrade the lifetime over a period of 8 hours, with nominal amplifier (TA, New Focus TA-7600). One reference laser is stabilized to a performance routinely restored after 16 hours of off time. temperature-controlled vapor cell module (Vescent D2-210) via saturated External to the physics package and coil housing is an aluminum absorption spectroscopy. The other two lasers are stabilized via frequency structure that provides mechanical stability for collimators, mirrors, and offset locks to the D cooling and repump transitions. In addition to the two cameras relative to the physics package. Each camera is a compact lasers’ internal isolators, a pigtailed isolator from Thorlabs provides another CMOS unit (Basler ACE acA2040-180 kmNIR) with 2048 × 2048 pixels, 30 dB of isolation from any light back-coupled downstream of the TA 5.5 μm pixel size, a quantum efficiency of 35%, and a dynamic range of input. Both TA outputs are distributed via fiber splitters and switches, 10 Bits. Both cameras are thermally strapped to the water cooling loop. We providing light for 2D- and 3D- MOTs, absorption imaging, and optical label the camera positioned to the side of the science cell that views the y– pumping. All frequency adjustments are made by controlling the relative z plane perpendicular to chip surface as the “side camera,” and the camera frequencies of the offset locks. While these TAs are capable of outputs up secured with imaging optics above a 3 mm circular window integrated into to 0.5 W, they are set to operate between 250 mW and 400 mW, for K and the chip surface as the “through-chip camera.” For a typical trapped cloud Rb, respectively. located a few hundred microns below the atom chip surface, the diameter of this window provides a potential numerical aperture up to 0.9. The MOT beam collimators are provided by Schäfter+ Kirchhoff and feature a Electronics titanium housing in addition to integrated clean up polarizers and quarter The presently reported results utilized ground support equipment (GSE) to wave plates. Four of these collimators produce 12 mm-diameter beams for power and control the flight science module, in parallel with the the 3D-MOT and side camera imaging, and two more collimators at the integration of CAL’s flight electronics and software. For GSE laser operation, level of the source cell produce 2D-MOT beams with an elliptical cross we use New Focus TLB-6700 tunable laser controllers, with Vescent section of 12 mm by 24 mm. Opposite the 2D- and 3D-MOT beam Photonics D2-135 offset phase lock servos and a D2-125 servo providing collimators are retro-reflecting mirrors, doubling the respective powers as the reference frequency lock. The TAs are each run by a New Focus TA- seen by the atoms. As shown in Fig. 4, the optical distribution system 7600 tampered amplifier controller. The drivers for both Leoni and Agiltron provides the same cooling and repump frequencies for each species to fiber switches are integrated within the switch assemblies. Currents to the both the 2D- and 3D-MOTs. A 4 mm diameter collimator at the bottom of magnet coils are sourced from Kepco bipolar power supplies, while npj Microgravity (2018) 16 Published in cooperation with the Biodesign Institute at Arizona State University, with the support of NASA NASA’s Cold Atom Lab (CAL): system development and ground ER Elliott et al. Fig. 4 Schematic of the CAL laser system for dual-species MOT operation, state preparation, and absorption imaging. Laser light is sourced from external cavity diode lasers (ECDL), with a reference laser for each species, labeled (a) for K and (d) for Rb, locked to an atomic line via frequency modulated spectroscopy (FMS). The potassium cooling (b) and repump (c) light is amplified in the same tapered amplifier (TA), where we have observed no intensity fluctuations from potential mode competition. For rubidium, only the cooling light (e) seeds the 780 nm TA, with the Rb repump laser output (f) propagating without amplification. The cooling and repump light for Rb and K is recombined and directed to both the 2D- and 3D-MOTs, while a switching assembly directs additional light to either the 2D-MOT push beam, the optical pumping path, or the imaging path. This fiber path selection is determined by high-extinction fiber coupled LEONI mechanical switches on the sub-ms timescale, while short pulse generation is accomplished with Agiltron Nanospeed electro-optical switches. All switches are fiber coupled. The n × m notation is [number of switch inputs] × [number of switch outputs]. The majority of our light losses occur through the switches, which are about 70% efficient. Per species, the 3D-MOT and 2D-MOT path provides up to 20 mW and 30 mW, respectively, at collimated outputs within the science module. The switch that distributes light to the imaging, optical pumping, and push beam, produces up to 2 mW of light from its output per species. This full power is used for optical pumping, and attenuated to 0.2–0.9 mW at the output paths for imaging and push beams. The flight system includes a seventh ECDL at 785 nm (not pictured) to be combined with a future, on-orbit upgrade to the physics package, providing the additional capability of Bragg interferometry Fig. 5 (a) Layout of the ground test equipment comprising the microwave and radio frequency (RF) chain for forced evaporation of Rb. Tunable RF signals are generated by a 80 MHz arbitrary waveform generator (AWG) and sent through a voltage-controlled attenuator and an amplifier. A tunable microwave source results from mixing a separate AWG and voltage-controlled attenuator with the output of a synthesized CW generator. We implement a fast and high-extinction ratio microwave switch followed by a chain of amplifiers. (b) Physics package science cell with loop antennas positioned millimeters above the atom chip assembly. The RF emitter is a 10 mm-diameter double loop of wire radiating 1–50 MHz, and the microwave antenna is a 12.5 mm-diameter single loop of wire operating within 100 MHz of 6.834 GHz. The capabilities of this ground test equipment allow us to set a fixed carrier wave at 6.846 GHz, the frequency resonant with the Rb |21⟩ to |11⟩ transition at a 8.4 G trap bottom to constantly remove unwanted |21⟩ atoms while evaporating uninterrupted with the upper side band. The CAL flight electronics, however, produce one pure tone. We therefore adopted our procedure to follow the capabilities of the flight instrument 87 39,41 currents for the atom chip traces come from ColdQuanta’s two-channel Dual-species Rb- K 2D MOT atomic source, 3D MOT, and atom chip driver. RF and microwave sources are shown in Fig. 5. Finally, atom chip transfer procedures experimental timing is accomplished with ColdQuanta's commercial Because complexity and size of the CAL instrument is reduced through the computer control system. CAL’s full flight system incorporates electronics use of an all fibered distribution system, digital switches, a single tapered designed and tested by JPL engineers specifically for ISS compatibility. amplifier for each species, and overlapped MOTs, our potassium cooling Published in cooperation with the Biodesign Institute at Arizona State University, with the support of NASA npj Microgravity (2018) 16 NASA’s Cold Atom Lab (CAL): system development and ground ER Elliott et al. strategy differs from other established techniques. We forego the sub- scientist for fundamental physics in NASA’s Division of Space Life and Physical doppler cooling methods that require higher laser power and variable, Sciences Research and Applications. This work was carried out at the Jet Propulsion 55,56 independent intensity control of each K beam, as well as the direct K Laboratory, California Institute of Technology, under a contract with the National evaporation methods that rely on an optical trap. Instead, the primary Aeronautics and Space Administration. CAL is supported by NASA’s Space Life and goal of our laser cooling procedure is to maximize the amount of rubidium Physical Sciences and The International Space Station Program Office. Government to sympathetically cool the available potassium. sponsorship is acknowledged. © 2018 All rights reserved. The prioritized laser cooling of Rb closely follows the recipe described in ref. with K frequencies adjusted to minimize losses and heating. Once the parameters for the simultaneous loading of Rb and K into the chip trap AUTHOR CONTRIBUTIONS were found, returning to the most efficient loading of Rb alone required D.C.A. led ground testbed planning, construction, development, and flight science only minor adjustments to bias fields. However, the absolute atom number module integration. D.C.A., E.R.E., M.C.K., and J.R.W. were responsible for hardware and temperature is more favorable when loading Rb alone (accomplished modification, daily experimental operation, optimization, and data acquisition. E.R.E. simply by shuttering the K light sources), owing mostly to the light induced analyzed the associated data and prepared the manuscript with M.C.K. R.J.T. collisions in the dual species MOT. We therefore summarize our loading of proposed the experiment and coordinated with principal investigators as CAL project the chip trap assuming the presence of both Rb and K. Generally, we also scientist. All authors read, edited, and approved the final manuscript. find that very similar powers and detunings from the respective cooling 41 39 and repump transitions work equally well for K and K. We begin dual species MOT loading by initially applying rubidium light ADDITIONAL INFORMATION only. Both dispensers in the source cell are continuously operated at Competing interests: The authors declare no competing interests. approximately 2.2–2.8 A, saturating the Rb 3D-MOT within 1 s. The dual species 2D-MOT achieves an atomic flux on the order of 10 atoms/s for 87 8 39 Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims Rb and 10 for K. In order to limit the heating due to light assisted 41 39 in published maps and institutional affiliations. collisions, we unshutter the K light to load Kor K on top of the rubidium during the last second of this loading phase. Our loading frequencies for Rb cooling and repump light are −1.5 Γ and −2.4 Γ, respectively, while the K cooling light operates at −4.8 Γ with the K REFERENCES repump light set at −3.4 Γ. After the dual-species MOT loading, we shutter 1. Kovachy, T. et al. Quantum superposition at the half-metre scale. Nature 528, the push beams and compress each MOT by increasing the field gradient 530–533 (2015). by a factor of 2.8 from 13 G/cm and simultaneously ramping the Rb 2. Geiger, R. et al. Matter-wave laser interferometric gravitation antenna (MIGA): cooling frequency from −4.8 Γ to −13.2 Γ over the course of 25 ms, while new perspectives for fundamental physics and geosciences. arXiv:1505.07137 the Rb repump light moves from its MOT loading value to −9.1 Γ. Here, (2015). the K cooling light also ramps slightly from −3.7 Γ to −3.2 Γ while its 3. Zhou, L. et al. Development of an atom gravimeter and status of the 10-meter repumper remains fixed at −3.3 Γ. Next, we effectively zero the magnetic atom interferometer for precision gravity measurement. General. Relativ. Gravit. field for 2.3 ms of polarization gradient cooling of Rb, further detuning 43, 1931 (2011). 87 87 the Rb cooling to −28 Γ while the Rb repump light is placed at −3.3 Γ. 4. Bidela, Y. et al. Compact cold atom gravimeter for field applications. Appl. Phys. Though we are unable to execute a parallel potassium sub-Doppler cooling Lett. 102, 144107 (2013). 55,56 protocol due to fixed optical intensity, we do observe a drop in the 5. Freier, C. et al. Mobile quantum gravity sensor with unprecedented stability. J. potassium temperature during this stage by sweeping its cooling light Phys. Conf. Ser. 723, 12050 (2016). from −2.5 Γ to −1.8 Γ and fixing its repump at −2.6 Γ. Lastly, over the 6. Lautier, J. et al. Hybridizing matter-wave and classical accelerometers. Appl. Phys. course of 2 ms, we optically pump Rb and K to their respective |22⟩ state Lett. 105, 144102 (2014). with 2 mW (per species) of circularly polarized light in preparation for the 7. Cold Quanta http://www.coldquanta.com/ (2017). remaining stages of magnetic trapping. This optical pumping stage 8. Muquans http://www.muquans.com/ (2017). increases the number of trapped atoms by roughly a factor of 2.5. After 9. AOSense http://aosense.com/ (2017). obtaining a polarized sample, all light is shuttered while the MOT coil fields 10. Muntinga, H. et al. Interferometry with Bose–Einstein condensates in micro- snap back on to form a magnetic quadrupole trap with a 100 G/cm field gravity. Phys. Rev. Lett. 110, 093602 (2013). gradient at a position approximately 20 mm below the atom chip surface. 11. van Zoest, T. et al. Bose–Einstein condensation in microgravity. Science 328, 1540 The atoms are then transported to the chip by turning on the second, (2010). vertically offset pair of transfer coils to 100 G/cm while simultaneously 12. Kulas, S. et al. Miniaturized lab system for future cold atom experiments in relaxing the quadrupole trap formed by the MOT coils. The large overlap of microgravity. Microgravity Sci. Technol. 29, 37 (2017). the MOT coils and the transfer coils minimizes heating during this 13. Stern, G. et al. Light-pulse atom interferometry in microgravity. Eur. Phys. J. D. 53, transport away from the MOT region. In contrast, our chip trap is non- 353–357 (2009). adiabatically loaded from the quadrupole of the transfer coils using a 14. Barret, B. et al. Dual matter-wave inertial sensors in weightlessness. Nat. Commun. throw and catch method. The transfer currents are chosen to match the 7, 13786 (2016). gradient that will be formed by the final chip trap as closely as possible, 15. Altenbuchner, L. et al. MORABA—overview on DLR’s mobile rocket base and before we immediately switch the transfer currents from a gradient projects. In SpaceOps 2012 Conference Proceedings (2012). configuration to a bias in the x-direction as fast as the inductance of the 16. Schkolnik, V. et al. A compact and robust diode laser system for atom inter- coils allows. This field is combined with bias fields in the y- and z- ferometry on a sounding rocket. Appl. Phys. B 122,1–8 (2016). directions, and atom chip currents of 3.2 A in the main Z-trace plus 0.362 A 17. Lezius, M. et al. Space-borne frequency comb metrology. Optica 3, 1381–1387 in the dimple trace, producing a trap with trapping frequencies of (ω , ω , (2016). x y ω ) = 2π × (207, 967, 1050) Hz for rubidium, (ω , ω , ω ) = 2π × (310, 1450, 18. Dinkelaker, A. N. et al. Autonomous frequency stabilization of two extended- z x y z 1570) for potassium, and a trap bottom of 8.4 G. cavity diode lasers at the potassium wavelength on a sounding rocket. Appl. Opt. 56, 1388–1396 (2017). 19. MAIUS 1—first Bose–Einstein Condensate Generated in Space. http://www.dlr.de/ Data availability dlr/en/desktopdefault.aspx/tabid-10081/151_read-20337/gallery/25194/ All relevant data are available from the authors. #/gallery/25194 (2017). 20. Elder, B. et al. Chameleon dark energy and atom interferometry. Phys. Rev. D. 94, 044051 (2016). ACKNOWLEDGEMENTS 21. Aguilera, D. N. et al. STE-QUEST-test of the universality of free fall using cold atom We gratefully acknowledge the support of CAL Project Manager Robert Shotwell, CAL interferometry. Class. Quantum Grav. 31, 115010 (2014). Deputy Project Manager Kamal Oudrhiri, JPL’s Communications Architectures & 22. Dimopoulos, S., Graham, P. W., Hogan, J. M. & Kasevich, M. A. Testing general Research Section Manager Norman Lay, JPL’s Quantum Sciences and Technology relativity with atom interferometry. Phys. Rev. Lett. 98, 111102 (2007). 23. Graham, P. W., Hogan, J. M., Kasevich, M. A. & Rajendran, S. New method for Group Supervisor Nan Yu, along with the entire CAL team. We thank Tyler Winn, Ken gravitational wave detection with atomic sensors. Phys. Rev. Lett. 110, 171102 Muse, James Kellogg, Dmitry Strekalov, James Kohel, Sascha Kulas, and Emmanuel (2013). Decrossas for technical support in the ground testbed. We also express our deepest appreciation for the long-term support and vision of Mark Lee, senior program npj Microgravity (2018) 16 Published in cooperation with the Biodesign Institute at Arizona State University, with the support of NASA NASA’s Cold Atom Lab (CAL): system development and ground ER Elliott et al. 24. Dimopoulos, S., Graham, P. W., Hogan, J. M., Kasevich, M. A. & Rajendran, S. 46. Roati, G. Quantum degenerate potassium–rubidium mixtures. Ph.D. thesis, Uni- Gravitational wave detection with atom interferometry. Phys. Lett. A 678,37 versitegli Studi di Trento (2003). (2009). 47. Bigelow, N. Consortium for Ultracold Atoms in Space. https://taskbook.nasaprs. 25. HYPER: Hyper precision atom interferometry in space, ESA CDF-09, 2000. com/Publication/index.cfm? 26. Hogan, J. M. et al. An atomic gravitational wave interferometric sensor in low action=public_query_taskbook_content&TASKID=10085 (2015). earth orbit (agis-leo). General. Relativ. Gravit. 43, 1953–2009 (2011). 48. Sackett, C. Development of Atom Interferometry Experiments for the International 27. Oi, D. K. L. et al. Nanosatellites for quantum science and technology. Contemp. Space Station’s Cold Atom Laboratory. https://taskbook.nasaprs.com/Publication/ Phys. 58,25–52 (2017). index.cfm?action=public_query_taskbook_content&TASKID=11097 (2017). 28. Schuldt, T., Schubert, C. & Krutzik, M. Design of a dual species atom inter- 49. Williams, J. Fundamental Interactions for Atom Interferometry with Ultracold ferometer for space. Exp. Astron. 39, 167–206 (2015). Quantum Gases in A Microgravity Environment. https://taskbook.nasaprs.com/ 29. Heß, M. et al. The ACES mission: system development and test status. Acta Publication/index.cfm?action=public_query_taskbook_content&TASKID=11101 Astronautica 69, 929–938 (2012). (2017). 30. Bongs, K. et al. Development of a strontium optical lattice clock for the soc 50. D’Incao, J. P., Krutzik, M., Elliott, E. & Williams, J. R. Enhanced association and mission on the ISS. Comptes Rendus Physique 16, 553–564 (2015). dissociation of heteronuclear feshbach molecules in a microgravity environment. 31. Tino, G. et al. Precision gravity tests with atom interferometry in space. Nucl. Phys. Phys. Rev. A 95, 012701 (2017). B-PS 243-244, 203–217 (2013). 51. Lundblad, N. Microgravity Dynamics of Bubble-Geometry Bose–Einstein Con- 32. Kaltenbaek, R. et al. Macroscopic quantum resonators (MAQRO). Exp. Astron. 34, densates. https://taskbook.nasaprs.com/Publication/index.cfm?action=public_ 123–164 (2012). query_taskbook_content&TASKID=11095 (2017). 33. Cacciapuoti, L. & Salomon, C. Atomic clock ensemble in space. J. Phys. Conf. Ser. 52. Cornell, E. Zero-G Studies of Few-body and Many-body Physics. https://taskbook. 327, 012049 (2011). nasaprs.com/Publication/index.cfm?action=public_query_taskbook_content& 34. Williams, J. R., Chiow, S.-W., Yu, N. & Müller, H. Quantum test of the equivalence TASKID=11096 (2017). principle and space-time aboard the international space station. New J. Phys. 18, 53. Dieckmann, K., Spreeuw, R. J. C., Weidemüller, M. & Walraven, J. T. M. Two- 025018 (2016). dimensional magneto-optical trap as a source of slow atoms. Phys. Rev. A 58, 35. Thompson, R. Science envelope requirements document (SERD) for cold atom 3891 (1998). laboratory. Tech. Rep. Jet Propulsion Laboratory (2013). 54. Salim, E. A., Caliga, S. C., Pfeiffer, J. B. & Anderson, D. Z. High resolution imaging 36. Farkas, D. M., Salim, E. A. & Ramirez-Serrano, J. Production of rubidium and optical control of Bose–Einstein condensates in an atom chip magnetic trap. Bose–Einstein condensates at a 1 Hz rate. arXiv:1403.4641v2 (2014). Appl. Phys. Lett. 102, 084104 (2013). 37. Silber, C. et al. Quantum-degenerate mixture of fermionic lithium and bosonic 55. Gokhroo, V., Rajalakshmi, G., Easwaran, R. K. & Unnikrishnan, C. S. Sub-doppler rubidium gases. Phys. Rev. Lett. 95, 170408 (2005). deep-cooled bosonic and fermionic isotopes of potassium in a compact 2D -3D 38. Wang, P., Xiong, D., Fu, Z. & Zhang, J. Experimental investigation of evaporative MOT set-up. J. Phys. B At. Mol. Opt. Phys. 44, 115307 (2011). 87 40 cooling mixtue of bosonic Rb and fermionic K atoms with microwave and 56. Landini, M. et al. Sub-doppler laser cooling of potassium atoms. Phys. Rev. A 84, radio frequency radiation. Chin. Phys. B 20, 016701 (2011). 043432 (2011). 39. Marzok, C., Deh, B., Courteille, P. & Zimmermann, C. Ultracold thermalization of Li and Rb. Phys. Rev. A 76, 052704 (2007). 40. Haas, M. et al. Species-selective microwave cooling of a mixture of rubidium and Open Access This article is licensed under a Creative Commons caesium atoms. New J. Phys. 9, 147 (2007). Attribution 4.0 International License, which permits use, sharing, 41. Taglieber, M., Voigt, A., Aoki, T., Hänsch, T. & Dieckmann, K. Quantum degenerate adaptation, distribution and reproduction in any medium or format, as long as you give two-species fermi–fermi mixture coexisting with a Bose–Einstein condensate. appropriate credit to the original author(s) and the source, provide a link to the Creative Phys. Rev. Lett. 100, 010401 (2008). Commons license, and indicate if changes were made. The images or other third party 42. Campbell, R. et al. Efficient production of large K Bose–Einstein condensates. material in this article are included in the article’s Creative Commons license, unless Phys. Rev. A 82, 063611 (2010). indicated otherwise in a credit line to the material. If material is not included in the 43. Modugno, G. et al. Bose–Einstein condensation of potassium atoms by sympa- article’s Creative Commons license and your intended use is not permitted by statutory thetic cooling. Science 294, 1320 (2001). regulation or exceeds the permitted use, you will need to obtain permission directly 44. Xiong, D., Wang, P., Chen, H. & Zhang, J. Species-selective microwave cooling of a from the copyright holder. To view a copy of this license, visit http://creativecommons. mixture of rubidium and caesium atoms. Chin. Opt. Lett. 8, 351 (2010). org/licenses/by/4.0/. 45. Modugno, G., Modugno, M., Riboli, F., Roati, G. & Inguscio, M. Two atomic species superfluid. Phys. Rev. Lett. 89, 170406 (2002). © The Author(s) 2018 Published in cooperation with the Biodesign Institute at Arizona State University, with the support of NASA npj Microgravity (2018) 16 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png npj Microgravity Springer Journals

NASA’s Cold Atom Lab (CAL): system development and ground test status

Download PDF
7 pages

Loading next page...
 
/lp/springer-journals/nasa-s-cold-atom-lab-cal-system-development-and-ground-test-status-QT93Gii3RE

References (55)

Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s)
Subject
Life Sciences; Life Sciences, general; Classical and Continuum Physics; Biotechnology; Immunology; Space Sciences (including Extraterrestrial Physics, Space Exploration and Astronautics) ; Applied Microbiology
eISSN
2373-8065
DOI
10.1038/s41526-018-0049-9
Publisher site
See Article on Publisher Site

Abstract

www.nature.com/npjmgrav ARTICLE OPEN NASA’s Cold Atom Lab (CAL): system development and ground test status 1 1,2 1 1 1 Ethan R. Elliott , Markus C. Krutzik , Jason R. Williams , Robert J. Thompson and David C. Aveline We report the status of the Cold Atom Lab (CAL) instrument to be operated aboard the International Space Station (ISS). Utilizing a 87 39 41 compact atom chip-based system to create ultracold mixtures and degenerate samples of Rb, K, and K, CAL is a multi-user facility developed by NASA’s Jet Propulsion Laboratory to provide the first persistent quantum gas platform in the microgravity conditions of space. Within this unique environment, atom traps can be decompressed to arbitrarily weak confining potentials, producing a new regime of picokelvin temperatures and ultra-low densities. Further, the complete removal of these confining potential allows the free fall evolution of ultracold clouds to be observed on unprecedented timescales compared to earthbound instruments. This unique facility will enable novel ultracold atom research to be remotely performed by an international group of principle investigators with broad applications in fundamental physics and inertial sensing. Here, we describe the development and validation of critical CAL technologies, including demonstration of the first on-chip Bose–Einstein condensation (BEC) of Rb with 39 87 41 87 microwave-based evaporation and the generation of ultracold dual-species quantum gas mixtures of K/ Rb and K/ Rb in an atom chip trap via sympathetic cooling. npj Microgravity (2018) 4:16 ; doi:10.1038/s41526-018-0049-9 INTRODUCTION The persistent free fall condition of low Earth orbit is recognized as the next natural destination for cold atom work to take The thermal, random motion of atomic gases fundamentally limits advantage of microgravity. Proposals include both satellite-based their free space measurement time and restricts their manipula- 21,25–28 missions and experiments to operate aboard the Interna- tion. Forty years of breakthroughs in the creation of ultracold 29–34 tional Space Station (ISS). The ISS becomes particularly quantum gases have resulted in standardized cooling techniques attractive when compared to a satellite mission both in terms of and technologies that are mature to the point of supporting large cost and the constant human presence, which enables instrument 1–3 4–6 scale dedicated facilities, portable field-ready devices, and modifications and upgrades. Leveraging these advantages, the 7–9 commercially offered components. In recent decades, national Cold Atom Lab (CAL) is designed to provide the first ultracold consortia and experimental groups seeking still colder tempera- quantum gas experiment aboard the ISS by utilizing an apparatus tures and extended measurement times have successfully pushed developed, assembled, and qualified by NASA’s Jet Propulsion ultracold atoms and related technologies into the free fall Laboratory (JPL). CAL is intended to operate as multi-user facility 10–12 13,14 conditions of drop towers, zero-g airplanes, and sub- for principal investigators to remotely study ultracold and 15–18 87 39 41 orbital launch vehicles. Here, two substantial benefits are quantum degenerate samples of Rb, K, and K, including realized. The first is access to a new parameter regime of quantum dual-species mixtures of Rb and K. In this article, we report the gases at picokelvin temperatures and ultra-low densities. Broadly, system development and test status of the CAL instrument, highlighting the successful production of dual species ultracold this is achieved by removing the asymmetric tilt that gravity gases in CAL’s ground testbed facility. introduces along one direction of the confining potential used to Conceptually, the CAL system consists of three primary trap and cool atoms. It is then possible, for example, to subsystems: the science module, electronics, and the laser and decompress these traps far beyond what is achievable on the optical distribution system. In order to support reliability, repair, ground. Secondly, the wave nature of ultracold atoms released integration, and replacement of components, all three subsystems from these traps is the basis for inertial sensing atom inter- leverage commercially offered components in an architecture ferometers, the precision of which scales as the square of their free intended to be compact, simple, and modular. This modular space evolution time. Recent milestones of these ruggedized, design has not only allowed parallel development of different autonomous devices operating in reduced gravity include the subsystems but also various versions of the same subsystem. In generation of Bose–Einstein condensates (BEC) and the demon- particular, CAL’s ground testbed currently supports multiple stration of matter-wave interferometry. These pathfinders physics packages, which are based on a commercially available underline both the technical feasibility and the future potential dual-cell vacuum chamber featuring an atom chip. CAL’s ground for high-precision fundamental physics applications including testbed has also developed two distinct “science modules,” the searches for dark energy, quantum tests of Einstein’s equiva- project designation for physics packages prepared specifically in 21,22 23,24 lence principle, and long-baseline gravity wave detectors. the flight configuration of robustly integrated optics, magnetic 1 2 Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91109, USA and Humboldt-Universität zu Berlin, Newtonstr. 15, 12489 Berlin, Germany Correspondence: David C. Aveline (david.c.aveline@jpl.nasa.gov) Received: 18 January 2018 Revised: 16 April 2018 Accepted: 30 May 2018 Published in cooperation with the Biodesign Institute at Arizona State University, with the support of NASA NASA’s Cold Atom Lab (CAL): system development and ground ER Elliott et al. procedure loads a limited number of K atoms into the chip trap and cannot afford the atom losses inherent to removing the hottest potassium atoms through a direct evaporative process. Instead, we evaporate only Rb, which in turn sympathetically cools trapped potassium. This cooling is accomplished specifically with a microwave evaporation scheme, transferring the most energetic |22⟩ Rb to the anti-trapped |11⟩ state, a state transition 37–45 that is separated by more than 6.8 GHz. Because the spacing 41 39 of the similar hyperfine states in K and K are 254 MHz and 462 MHz, respectively, microwave signals at 6.8 GHz address only Rb. This is in contrast to a more typical RF evaporation scheme, which instead operates on states with frequency spacings that are almost exactly the same in Rb and either species of bosonic potassium, specifically transitioning trapped |22⟩ atoms to the more weakly trapped |21⟩ state before arriving at the non- magnetic |20⟩ state or anti-trapped |2−1⟩ state. However, it follows that microwave evaporation yields the comparative disadvantage of not immediately removing any |21⟩ Rb atoms, should they also exist in the trap. The source of the |21⟩ impurity during microwave evaporation of |22⟩ Rb has been suggested to 37,39 result from either inelastic spin-changing collisions or an ejected |11⟩ atom moving to a point of higher magnetic field where the microwave knife becomes resonant with the |11⟩ to | 38,40,44 21⟩ transition. Consistent with either of these possible cases is that the |21⟩ impurity is observed to emerge during the evaporation process, and not simply as the result of inefficient optical pumping. If not addressed, these |21⟩ atoms can act as an additional heat load to be sympathetically cooled or potentially remove potassium atoms via energy released from interspecies collisions that do not conserve electronic spin states. Our microwave evaporation protocol incorporates five linear fre- quency ramps that at low temperatures are interspersed with brief clearing stages at a second fixed frequency (6.846 GHz) 87 39 Fig. 1 False color absorption images of Rb and K after 3 ms of resonant with the Rb |21⟩ to |11⟩ transition at the 8.4 G trap expansion from an atom chip trap at five different stages of bottom to recurrently remove the unwanted |21⟩ atoms. Begin- microwave-induced evaporation of rubidium and the corresponding ning 60 MHz above the bare hyperfine transition at 6.834 GHz, sympathetic cooling of potassium. Higher temperature/atom these five ramps together last a total of 1.5 s, with up to one number data are shown for illustrative purposes. The K atoms hundred equally spaced jumps to 6.846 GHz. loaded onto the chip are initially so diffuse that our imaging fails to Following evaporation, we decompress the trap via a 70% detect them. The last stage shows 70,000 K atoms at a reduction of the the bias fields in 100 ms before release and temperature of 1.6 μK and phase space density (PSD) of 0.05. Rb 39 87 numbers in the final stage are 155,000 atoms at a temperature of imaging. Initial results of sympathetically cooling K with Rb 39 4 1.6 μK and a PSD of 0.025 yield K temperatures of 170 nK with 1.1 × 10 atoms and a phase space density (PSD) of 0.6. For illustrative purposes, Fig. 1 shows a coils, two CMOS cameras, microwave and radio frequency (RF) higher temperature/atom number example through five stages of emitters, a water-based cooling loop, and magnetic shields. evaporation. Additionally, our system has cooled K to degen- Details of the science module and physics package can be found eracy, with current results yielding an 11% condensate fraction in the “Materials and Methods” section of this paper. This modular out of 4200 total atoms at 76 nK and a PSD of 3.5. The larger 41 87 39 87 architecture also allows more readily available prototype/ground collisional cross section of K/ Rb compared to K/ Rb support systems of optics and electronics to test the functionality increases not only the comparative evaporation efficiency, but of the science modules during development of the flight models. also yields more effective sympathetic cooling during the initial Specifically, all degenerate and ultracold atom results reported in loading of the magnetic quadrupole trap. Repeating the above this article use the flight science module, flight equivalent lasers/ procedure without loading potassium produces Rb BECs with optical fibers, and ground support electronics as the flight 3×10 total atoms and a condensate fraction of 10%, comparable electronics were finalized. The electronics described in Materials to the performance of our Rb BECs formed using standard RF and Methods reflect these ground versions. evaporation (5 × 10 atoms with 10% BEC fraction). To the best of our knowledge, these results mark the first use of an atom chip to evaporate Rb to condensation using microwaves, sympatheti- RESULTS cally cool either bosonic potassium isotope, and produce We have successfully produced degenerate samples of Rb and degenerate K. 41 39 K in addition to nearly degenerate K in the atom chip While the first group to demonstrate interspecies sympathetic magnetic trap within CAL’s flight science module. Beginning with cooling to a BEC did not include techniques to directly remove | 87 39 87 43 the loading of the atom chip trap from dual species Rb/ Kor 21⟩ Rb atoms, their subsequent implementation of a counter- 87 41 Rb/ K in 3D and 2D magneto-optical traps (MOTs), the measure against the |21⟩ state increased evaporation efficiency to 45,46 production of a Rb BEC follows using either RF or microwave- allow the first mixture of dual species BECs. All later induced evaporation though the selective rejection of the most implementations of Rb microwave evaporation have incorpo- energetic Rb atoms via transfer from the low magnetic field- rated an additional frequency targeting the |21⟩ atoms through- 38– seeking |F = 2, m = 2⟩ state to a high-field-seeking, anti-trapped out the entire evaporation ramp, either continuously operating 45 37 state. As described in the section “Materials and Methods,” our or repeatedly pulsed. As noted, CAL employs a variation of the npj Microgravity (2018) 16 Published in cooperation with the Biodesign Institute at Arizona State University, with the support of NASA 1234567890():,; NASA’s Cold Atom Lab (CAL): system development and ground ER Elliott et al. Fig. 2 Necessity of two frequencies to remove the |21⟩ impurity during microwave evaporation of |22⟩ Rb to BEC in a harmonic magnetic trap with trap bottom B . If a |21⟩ population exists in thermal equilibrium with |22⟩ atoms, it will form a cloud with a larger mean-square cloud size than the |22⟩ cloud due to weaker coupling to the magnetic field. (a) Maximum total energy of a trapped atom in either state exceeds μ B , where μ is the Bohr magneton, and a single frequency (green arrow) that evaporates the most energetic |22⟩ atoms to the anti- B 0 B trapped |11⟩ state will also remove the hottest |21⟩ atoms. (b) The maximum total energy is less than μ B , and the single |22⟩ evaporation B 0 frequency is too high to affect the |21⟩ atoms. (c) A second frequency (yellow arrow) resonant with the |21⟩ to |11⟩ transition at B removes all trapped 21 atoms latter technique. However, at higher temperatures, we find it is resonances to control differential center-of-mass distributions of 49,50 possible to evaporate sufficiently with only a single frequency. We dual-species quantum gas mixtures, bubble-shell geometries note that this is due to a combination of two effects that exist for | for Bose–Einstein condensates, as well as few-body systems in 21⟩ and |22⟩ atoms simultaneously confined in a harmonic new temperature and density regimes that are prerequisites for magnetic trap with trap bottom B . First, a |21⟩ population is more the next generation of Efimov experiments. After the conclusion weakly coupled to the magnetic field, experiences a looser of these experiments, the instrument will either enter a confinement, and when thermal equilibrium is assumed, forms a decommissioning phase or, through the use of orbital replace- larger cloud compared to a |22⟩ population. Secondly, a |21⟩ atom ment units, receive novel, upgraded capabilities for extended at the same magnetic field as a |22⟩ atom requires a lower mission operation and new microgravity experiments of advan- frequency microwave field to evaporate to the |11⟩ state. cing scope and complexity. Quantitatively, it is easily shown for this case that if a |22⟩ atom with total energy E relative to its trap bottom can climb the trap MATERIALS AND METHODS wall to a magnetic field of B, related by E = m g μ (B − B ) where F L B 0 μ is the Bohr magneton and g is the Landé g-factor, then a |21⟩ Science module B L atom with the same total energy E can reach a higher magnetic The physics package resides inside CAL’s science module, which also field given by 2B − B . If the most energetic |22⟩ atoms reach B,a houses all the hardware that maintains the appropriate local environment of magnetic fields, electromagnetic fields (optical, RF and microwave), as microwave field with frequency ω + 3Bμ /(2ℏ), where ω /2π is hf B hf well as thermal and mechanical stability (Fig. 3). Forming the outer the zero field Rb hyperfine splitting, is required to transition boundary of the science module is a dual layer magnetic shield, which them to the |11⟩ state. Comparison to the frequency that mitigates the significant and regularly modulated external magnetic fields evaporates the most energetic |21⟩ atoms, given by ω + (2B − hf experienced on the ISS. A bulkhead on the side of shield holds all B )μ /ℏ, yields a critical microwave frequency, ω = ω + 3B μ /ℏ, 0 B c hf 0 B necessary optical, electrical and thermal feedthroughs for the physics above which a single frequency is sufficient to evaporate |22⟩ package, including the water cooling loop, seven optical fibers (780 nm atoms while still decreasing the |21⟩ population, but below which cooling, 780 nm repump, 767 nm cooling and repump, imaging, optical a second frequency becomes necessary to remove the remaining | pumping, 2D-MOT push, and Bragg beam), power for the magnetic coils, 21⟩ impurity (Fig. 2). Similarly, there is also a critical total energy of atom chip currents, alkali metal dispensers, ion pump, two cameras (plus a trapped atom in either state relative to its respective trap datalinks), and RF/microwave signals. CAL’s physics package (Fig. 3a) is based on ColdQuanta’s commercially bottom given by μ B . Following this simplified model, a harmonic B 0 7,36 offered RuBECi vacuum chamber, modified for flight and CAL-specific magnetic trap with a higher trap bottom would therefore have a science objectives. These modifications include a second atomic source for relatively larger region where two frequencies are necessary for potassium, specialized anti-reflective coatings, a unique configuration of efficient microwave evaporation. atom chip conductive paths, as well as features for improved electrical, thermal, and mechanical integrity. The physics package consists of a dual glass-cell vacuum chamber, with one cell (source cell) housing two 12 mm DISCUSSION alkali metal dispensers (AMD, SAES) and a non-evaporable getter (NEG, Following the work presented here, one science module will SAES ST175), while the second (science cell) holds an ultra-high vacuum remain at JPL facilities for PI-specific experiment development and with an atom chip forming its ceiling. In order to maintain pressures below −10 10 Torr in the science cell, a miniaturized 2 L/s ion pump and a second further optimization of dual species quantum gas production, NEG are implemented in a six-flange stainless steel vacuum intersection. while the primary science module is integrated into the fight The source cell is surrounded by four rods lined with permanent magnets instrument for acceptance testing and final ground verification as for the production of overlapped dual species 2D-MOTs. A smooth silicon an autonomous instrument before delivery to the launch site. chip with a 0.75 mm diameter pinhole provides differential pumping with Following CAL’s arrival at the ISS, there will be an initial installation the science cell. This pinhole also allows the propagation of cold atoms phase by the onboard crew, including its first power up and basic from the 2D-MOT in the source cell to the 3D-MOT in the science cell. A communication tests between the instrument and the Ground push beam is directed along the axis of this pinhole from below the source Data System (GDS). In the subsequent commissioning phase, the cell, while partial reflection of this beam’s edges from the silicon surface 36,53 flight system performance will be tested and optimized for additionally boosts the loading rate of the 3D-MOT. The CAL science cell is surrounded by ten rectangular-shaped magnetic microgravity with the option of data downlinked to the GDS in coils, housed in an anodized aluminum structure that couples to the water real time. In the nominal operation phase, the CAL instrument will cooling loop via eight thermal straps. These enclosed coils produce the be run with a duty cycle of 1 min for up to 8 h per day during crew fields necessary for the MOT, magnetic transport, transfer into a chip trap, sleep, to minimize vibrations. These operations will be conducted and tuning near Feshbach resonances. The coils are capable of generating through a collaboration of the CAL PIs and the CAL team based at magnetic bias fields along each Cartesian axis. Along the x-axis, JPL, with each PI allocated schedule for their experiments to be independent coil control allows anti-Helmholtz configurations for mag- run. The proposed research will investigate applications of atom netic trapping gradients. Two separate pairs of vertically offset x-coils interferometry for future precision measurements in the areas of translate the zero-field position from the MOT location to the atom chip. 47,48 Earth observation and fundamental physics, Feshbach Incorporated into these upper x-coils (the transfer coils) are two turns of Published in cooperation with the Biodesign Institute at Arizona State University, with the support of NASA npj Microgravity (2018) 16 NASA’s Cold Atom Lab (CAL): system development and ground ER Elliott et al. Fig. 3 Layered overview of the CAL science module. (a) The physics package and magnetic coil assembly. MOT coils are shown in red, with the central axis of coils oriented along the x-axis. Transfer coils (in green) are vertically offset from the MOT coils. (b) Aluminum enclosure for the physics package securing the cameras and optical fiber collimators. Collimators and mirrors are emphasized in order to illustrate beam geometry, with red/orange arrows indicating effective directions of beam propagation. Two sets of 3D-MOT beams propagate diagonally in the y–z plane, while the third propagates along the x-axis through the MOT coils. Input for the optical pumping beam path (orange arrow) is positioned above the x-axis MOT beam collimator. The collimators opposite the two cameras provide light for absorption imaging, while the collimator below the through-chip camera also provides the push beam. The remaining two collimators surrounding the source cell send light to the 2D-MOT. Opposite the 2D- and 3D-MOT beam collimators are retro-reflecting mirrors (the retro-reflecting mirror for the x-axis MOT beam is not visible from this orientation). (c) The outside of this aluminum structure is bracketed to the water cooling loop and further enclosed by the magnetic shields with feed through connections from the other subsystems of the CAL instrument. Fully assembled, the dimensions are 46 cm × 30.5 cm × 58.5 cm, at a mass of 45 kg wires designated as the fast Feshbach coils (FFC). In conjunction with the the source cell directs a vertical beam through the center of the physics package to serve as either the push beam or the absorption imaging beam transfer coils in Helmholtz configuration (producing up to 325 G), the FFC for the through-chip camera. In order to provide efficient optical pumping, ramps the x-bias field by 3G within 50 μs. The upper side of the coil an additional dedicated output coupler and quarter wave plate on the side housing attaches to a breakout board for electrical connections to the of the science cell produces a slightly diverging beam that is roughly 1 cm atom chip, a microwave antenna, and a RF antenna. This breakout board in diameter at the location of the atoms. also concentrically secures these loop-style antennas a few mm above the atom chip (Fig. 5). Within the science cell, we have obtained a 10 s lifetime in the tight trap Laser and optical distribution system used for forced evaporation, where (ω , ω , ω )= 2π × (300, 1450, 1490) and x y z The optical distribution system of CAL, on the ground and in flight, is a variable hold time is introduced following three evaporation stages. A based on commercially available lasers and components to create an all variable hold time in the same trap following BEC production yields a optical fiber-based distribution system (Fig. 4). We operate one laser condensate lifetime of 2 s. Limitations to the trap lifetime result from noise system for both bosonic potassium isotopes at 766.701 nm, and an on the current drivers sourcing the currents for the magnetic trap (both the additional system for Rb at 780.24 nm. Laser cooling light is sourced by chip traces and coils) and vacuum quality. The flight electronics exceed the external cavity diode lasers (New Focus Vortex Plus) with a ruggedized performance of ground-use commercial chip current drivers, as the ground design for flight. Each laser features an internal 30 dB optical isolator and chip drivers achieve a lifetime of 8 s in the same configuration. We find that pigtailed optical fiber output coupler. The output light is guided through the lifetime is strongly influenced by the operation of the dispensers within single-mode polarization maintaining fibers (Corning PM 850) and fiber the source cell, although any related change in vacuum quality is below the splitter arrays from Evanescent Optics. range of the ion pump gauge. Continuous operation of both dispensers will For each atomic element, we use three separate lasers and a tapered gradually degrade the lifetime over a period of 8 hours, with nominal amplifier (TA, New Focus TA-7600). One reference laser is stabilized to a performance routinely restored after 16 hours of off time. temperature-controlled vapor cell module (Vescent D2-210) via saturated External to the physics package and coil housing is an aluminum absorption spectroscopy. The other two lasers are stabilized via frequency structure that provides mechanical stability for collimators, mirrors, and offset locks to the D cooling and repump transitions. In addition to the two cameras relative to the physics package. Each camera is a compact lasers’ internal isolators, a pigtailed isolator from Thorlabs provides another CMOS unit (Basler ACE acA2040-180 kmNIR) with 2048 × 2048 pixels, 30 dB of isolation from any light back-coupled downstream of the TA 5.5 μm pixel size, a quantum efficiency of 35%, and a dynamic range of input. Both TA outputs are distributed via fiber splitters and switches, 10 Bits. Both cameras are thermally strapped to the water cooling loop. We providing light for 2D- and 3D- MOTs, absorption imaging, and optical label the camera positioned to the side of the science cell that views the y– pumping. All frequency adjustments are made by controlling the relative z plane perpendicular to chip surface as the “side camera,” and the camera frequencies of the offset locks. While these TAs are capable of outputs up secured with imaging optics above a 3 mm circular window integrated into to 0.5 W, they are set to operate between 250 mW and 400 mW, for K and the chip surface as the “through-chip camera.” For a typical trapped cloud Rb, respectively. located a few hundred microns below the atom chip surface, the diameter of this window provides a potential numerical aperture up to 0.9. The MOT beam collimators are provided by Schäfter+ Kirchhoff and feature a Electronics titanium housing in addition to integrated clean up polarizers and quarter The presently reported results utilized ground support equipment (GSE) to wave plates. Four of these collimators produce 12 mm-diameter beams for power and control the flight science module, in parallel with the the 3D-MOT and side camera imaging, and two more collimators at the integration of CAL’s flight electronics and software. For GSE laser operation, level of the source cell produce 2D-MOT beams with an elliptical cross we use New Focus TLB-6700 tunable laser controllers, with Vescent section of 12 mm by 24 mm. Opposite the 2D- and 3D-MOT beam Photonics D2-135 offset phase lock servos and a D2-125 servo providing collimators are retro-reflecting mirrors, doubling the respective powers as the reference frequency lock. The TAs are each run by a New Focus TA- seen by the atoms. As shown in Fig. 4, the optical distribution system 7600 tampered amplifier controller. The drivers for both Leoni and Agiltron provides the same cooling and repump frequencies for each species to fiber switches are integrated within the switch assemblies. Currents to the both the 2D- and 3D-MOTs. A 4 mm diameter collimator at the bottom of magnet coils are sourced from Kepco bipolar power supplies, while npj Microgravity (2018) 16 Published in cooperation with the Biodesign Institute at Arizona State University, with the support of NASA NASA’s Cold Atom Lab (CAL): system development and ground ER Elliott et al. Fig. 4 Schematic of the CAL laser system for dual-species MOT operation, state preparation, and absorption imaging. Laser light is sourced from external cavity diode lasers (ECDL), with a reference laser for each species, labeled (a) for K and (d) for Rb, locked to an atomic line via frequency modulated spectroscopy (FMS). The potassium cooling (b) and repump (c) light is amplified in the same tapered amplifier (TA), where we have observed no intensity fluctuations from potential mode competition. For rubidium, only the cooling light (e) seeds the 780 nm TA, with the Rb repump laser output (f) propagating without amplification. The cooling and repump light for Rb and K is recombined and directed to both the 2D- and 3D-MOTs, while a switching assembly directs additional light to either the 2D-MOT push beam, the optical pumping path, or the imaging path. This fiber path selection is determined by high-extinction fiber coupled LEONI mechanical switches on the sub-ms timescale, while short pulse generation is accomplished with Agiltron Nanospeed electro-optical switches. All switches are fiber coupled. The n × m notation is [number of switch inputs] × [number of switch outputs]. The majority of our light losses occur through the switches, which are about 70% efficient. Per species, the 3D-MOT and 2D-MOT path provides up to 20 mW and 30 mW, respectively, at collimated outputs within the science module. The switch that distributes light to the imaging, optical pumping, and push beam, produces up to 2 mW of light from its output per species. This full power is used for optical pumping, and attenuated to 0.2–0.9 mW at the output paths for imaging and push beams. The flight system includes a seventh ECDL at 785 nm (not pictured) to be combined with a future, on-orbit upgrade to the physics package, providing the additional capability of Bragg interferometry Fig. 5 (a) Layout of the ground test equipment comprising the microwave and radio frequency (RF) chain for forced evaporation of Rb. Tunable RF signals are generated by a 80 MHz arbitrary waveform generator (AWG) and sent through a voltage-controlled attenuator and an amplifier. A tunable microwave source results from mixing a separate AWG and voltage-controlled attenuator with the output of a synthesized CW generator. We implement a fast and high-extinction ratio microwave switch followed by a chain of amplifiers. (b) Physics package science cell with loop antennas positioned millimeters above the atom chip assembly. The RF emitter is a 10 mm-diameter double loop of wire radiating 1–50 MHz, and the microwave antenna is a 12.5 mm-diameter single loop of wire operating within 100 MHz of 6.834 GHz. The capabilities of this ground test equipment allow us to set a fixed carrier wave at 6.846 GHz, the frequency resonant with the Rb |21⟩ to |11⟩ transition at a 8.4 G trap bottom to constantly remove unwanted |21⟩ atoms while evaporating uninterrupted with the upper side band. The CAL flight electronics, however, produce one pure tone. We therefore adopted our procedure to follow the capabilities of the flight instrument 87 39,41 currents for the atom chip traces come from ColdQuanta’s two-channel Dual-species Rb- K 2D MOT atomic source, 3D MOT, and atom chip driver. RF and microwave sources are shown in Fig. 5. Finally, atom chip transfer procedures experimental timing is accomplished with ColdQuanta's commercial Because complexity and size of the CAL instrument is reduced through the computer control system. CAL’s full flight system incorporates electronics use of an all fibered distribution system, digital switches, a single tapered designed and tested by JPL engineers specifically for ISS compatibility. amplifier for each species, and overlapped MOTs, our potassium cooling Published in cooperation with the Biodesign Institute at Arizona State University, with the support of NASA npj Microgravity (2018) 16 NASA’s Cold Atom Lab (CAL): system development and ground ER Elliott et al. strategy differs from other established techniques. We forego the sub- scientist for fundamental physics in NASA’s Division of Space Life and Physical doppler cooling methods that require higher laser power and variable, Sciences Research and Applications. This work was carried out at the Jet Propulsion 55,56 independent intensity control of each K beam, as well as the direct K Laboratory, California Institute of Technology, under a contract with the National evaporation methods that rely on an optical trap. Instead, the primary Aeronautics and Space Administration. CAL is supported by NASA’s Space Life and goal of our laser cooling procedure is to maximize the amount of rubidium Physical Sciences and The International Space Station Program Office. Government to sympathetically cool the available potassium. sponsorship is acknowledged. © 2018 All rights reserved. The prioritized laser cooling of Rb closely follows the recipe described in ref. with K frequencies adjusted to minimize losses and heating. Once the parameters for the simultaneous loading of Rb and K into the chip trap AUTHOR CONTRIBUTIONS were found, returning to the most efficient loading of Rb alone required D.C.A. led ground testbed planning, construction, development, and flight science only minor adjustments to bias fields. However, the absolute atom number module integration. D.C.A., E.R.E., M.C.K., and J.R.W. were responsible for hardware and temperature is more favorable when loading Rb alone (accomplished modification, daily experimental operation, optimization, and data acquisition. E.R.E. simply by shuttering the K light sources), owing mostly to the light induced analyzed the associated data and prepared the manuscript with M.C.K. R.J.T. collisions in the dual species MOT. We therefore summarize our loading of proposed the experiment and coordinated with principal investigators as CAL project the chip trap assuming the presence of both Rb and K. Generally, we also scientist. All authors read, edited, and approved the final manuscript. find that very similar powers and detunings from the respective cooling 41 39 and repump transitions work equally well for K and K. We begin dual species MOT loading by initially applying rubidium light ADDITIONAL INFORMATION only. Both dispensers in the source cell are continuously operated at Competing interests: The authors declare no competing interests. approximately 2.2–2.8 A, saturating the Rb 3D-MOT within 1 s. The dual species 2D-MOT achieves an atomic flux on the order of 10 atoms/s for 87 8 39 Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims Rb and 10 for K. In order to limit the heating due to light assisted 41 39 in published maps and institutional affiliations. collisions, we unshutter the K light to load Kor K on top of the rubidium during the last second of this loading phase. Our loading frequencies for Rb cooling and repump light are −1.5 Γ and −2.4 Γ, respectively, while the K cooling light operates at −4.8 Γ with the K REFERENCES repump light set at −3.4 Γ. After the dual-species MOT loading, we shutter 1. Kovachy, T. et al. Quantum superposition at the half-metre scale. Nature 528, the push beams and compress each MOT by increasing the field gradient 530–533 (2015). by a factor of 2.8 from 13 G/cm and simultaneously ramping the Rb 2. Geiger, R. et al. Matter-wave laser interferometric gravitation antenna (MIGA): cooling frequency from −4.8 Γ to −13.2 Γ over the course of 25 ms, while new perspectives for fundamental physics and geosciences. arXiv:1505.07137 the Rb repump light moves from its MOT loading value to −9.1 Γ. Here, (2015). the K cooling light also ramps slightly from −3.7 Γ to −3.2 Γ while its 3. Zhou, L. et al. Development of an atom gravimeter and status of the 10-meter repumper remains fixed at −3.3 Γ. Next, we effectively zero the magnetic atom interferometer for precision gravity measurement. General. Relativ. Gravit. field for 2.3 ms of polarization gradient cooling of Rb, further detuning 43, 1931 (2011). 87 87 the Rb cooling to −28 Γ while the Rb repump light is placed at −3.3 Γ. 4. Bidela, Y. et al. Compact cold atom gravimeter for field applications. Appl. Phys. Though we are unable to execute a parallel potassium sub-Doppler cooling Lett. 102, 144107 (2013). 55,56 protocol due to fixed optical intensity, we do observe a drop in the 5. Freier, C. et al. Mobile quantum gravity sensor with unprecedented stability. J. potassium temperature during this stage by sweeping its cooling light Phys. Conf. Ser. 723, 12050 (2016). from −2.5 Γ to −1.8 Γ and fixing its repump at −2.6 Γ. Lastly, over the 6. Lautier, J. et al. Hybridizing matter-wave and classical accelerometers. Appl. Phys. course of 2 ms, we optically pump Rb and K to their respective |22⟩ state Lett. 105, 144102 (2014). with 2 mW (per species) of circularly polarized light in preparation for the 7. Cold Quanta http://www.coldquanta.com/ (2017). remaining stages of magnetic trapping. This optical pumping stage 8. Muquans http://www.muquans.com/ (2017). increases the number of trapped atoms by roughly a factor of 2.5. After 9. AOSense http://aosense.com/ (2017). obtaining a polarized sample, all light is shuttered while the MOT coil fields 10. Muntinga, H. et al. Interferometry with Bose–Einstein condensates in micro- snap back on to form a magnetic quadrupole trap with a 100 G/cm field gravity. Phys. Rev. Lett. 110, 093602 (2013). gradient at a position approximately 20 mm below the atom chip surface. 11. van Zoest, T. et al. Bose–Einstein condensation in microgravity. Science 328, 1540 The atoms are then transported to the chip by turning on the second, (2010). vertically offset pair of transfer coils to 100 G/cm while simultaneously 12. Kulas, S. et al. Miniaturized lab system for future cold atom experiments in relaxing the quadrupole trap formed by the MOT coils. The large overlap of microgravity. Microgravity Sci. Technol. 29, 37 (2017). the MOT coils and the transfer coils minimizes heating during this 13. Stern, G. et al. Light-pulse atom interferometry in microgravity. Eur. Phys. J. D. 53, transport away from the MOT region. In contrast, our chip trap is non- 353–357 (2009). adiabatically loaded from the quadrupole of the transfer coils using a 14. Barret, B. et al. Dual matter-wave inertial sensors in weightlessness. Nat. Commun. throw and catch method. The transfer currents are chosen to match the 7, 13786 (2016). gradient that will be formed by the final chip trap as closely as possible, 15. Altenbuchner, L. et al. MORABA—overview on DLR’s mobile rocket base and before we immediately switch the transfer currents from a gradient projects. In SpaceOps 2012 Conference Proceedings (2012). configuration to a bias in the x-direction as fast as the inductance of the 16. Schkolnik, V. et al. A compact and robust diode laser system for atom inter- coils allows. This field is combined with bias fields in the y- and z- ferometry on a sounding rocket. Appl. Phys. B 122,1–8 (2016). directions, and atom chip currents of 3.2 A in the main Z-trace plus 0.362 A 17. Lezius, M. et al. Space-borne frequency comb metrology. Optica 3, 1381–1387 in the dimple trace, producing a trap with trapping frequencies of (ω , ω , (2016). x y ω ) = 2π × (207, 967, 1050) Hz for rubidium, (ω , ω , ω ) = 2π × (310, 1450, 18. Dinkelaker, A. N. et al. Autonomous frequency stabilization of two extended- z x y z 1570) for potassium, and a trap bottom of 8.4 G. cavity diode lasers at the potassium wavelength on a sounding rocket. Appl. Opt. 56, 1388–1396 (2017). 19. MAIUS 1—first Bose–Einstein Condensate Generated in Space. http://www.dlr.de/ Data availability dlr/en/desktopdefault.aspx/tabid-10081/151_read-20337/gallery/25194/ All relevant data are available from the authors. #/gallery/25194 (2017). 20. Elder, B. et al. Chameleon dark energy and atom interferometry. Phys. Rev. D. 94, 044051 (2016). ACKNOWLEDGEMENTS 21. Aguilera, D. N. et al. STE-QUEST-test of the universality of free fall using cold atom We gratefully acknowledge the support of CAL Project Manager Robert Shotwell, CAL interferometry. Class. Quantum Grav. 31, 115010 (2014). Deputy Project Manager Kamal Oudrhiri, JPL’s Communications Architectures & 22. Dimopoulos, S., Graham, P. W., Hogan, J. M. & Kasevich, M. A. Testing general Research Section Manager Norman Lay, JPL’s Quantum Sciences and Technology relativity with atom interferometry. Phys. Rev. Lett. 98, 111102 (2007). 23. Graham, P. W., Hogan, J. M., Kasevich, M. A. & Rajendran, S. New method for Group Supervisor Nan Yu, along with the entire CAL team. We thank Tyler Winn, Ken gravitational wave detection with atomic sensors. Phys. Rev. Lett. 110, 171102 Muse, James Kellogg, Dmitry Strekalov, James Kohel, Sascha Kulas, and Emmanuel (2013). Decrossas for technical support in the ground testbed. We also express our deepest appreciation for the long-term support and vision of Mark Lee, senior program npj Microgravity (2018) 16 Published in cooperation with the Biodesign Institute at Arizona State University, with the support of NASA NASA’s Cold Atom Lab (CAL): system development and ground ER Elliott et al. 24. Dimopoulos, S., Graham, P. W., Hogan, J. M., Kasevich, M. A. & Rajendran, S. 46. Roati, G. Quantum degenerate potassium–rubidium mixtures. Ph.D. thesis, Uni- Gravitational wave detection with atom interferometry. Phys. Lett. A 678,37 versitegli Studi di Trento (2003). (2009). 47. Bigelow, N. Consortium for Ultracold Atoms in Space. https://taskbook.nasaprs. 25. HYPER: Hyper precision atom interferometry in space, ESA CDF-09, 2000. com/Publication/index.cfm? 26. Hogan, J. M. et al. An atomic gravitational wave interferometric sensor in low action=public_query_taskbook_content&TASKID=10085 (2015). earth orbit (agis-leo). General. Relativ. Gravit. 43, 1953–2009 (2011). 48. Sackett, C. Development of Atom Interferometry Experiments for the International 27. Oi, D. K. L. et al. Nanosatellites for quantum science and technology. Contemp. Space Station’s Cold Atom Laboratory. https://taskbook.nasaprs.com/Publication/ Phys. 58,25–52 (2017). index.cfm?action=public_query_taskbook_content&TASKID=11097 (2017). 28. Schuldt, T., Schubert, C. & Krutzik, M. Design of a dual species atom inter- 49. Williams, J. Fundamental Interactions for Atom Interferometry with Ultracold ferometer for space. Exp. Astron. 39, 167–206 (2015). Quantum Gases in A Microgravity Environment. https://taskbook.nasaprs.com/ 29. Heß, M. et al. The ACES mission: system development and test status. Acta Publication/index.cfm?action=public_query_taskbook_content&TASKID=11101 Astronautica 69, 929–938 (2012). (2017). 30. Bongs, K. et al. Development of a strontium optical lattice clock for the soc 50. D’Incao, J. P., Krutzik, M., Elliott, E. & Williams, J. R. Enhanced association and mission on the ISS. Comptes Rendus Physique 16, 553–564 (2015). dissociation of heteronuclear feshbach molecules in a microgravity environment. 31. Tino, G. et al. Precision gravity tests with atom interferometry in space. Nucl. Phys. Phys. Rev. A 95, 012701 (2017). B-PS 243-244, 203–217 (2013). 51. Lundblad, N. Microgravity Dynamics of Bubble-Geometry Bose–Einstein Con- 32. Kaltenbaek, R. et al. Macroscopic quantum resonators (MAQRO). Exp. Astron. 34, densates. https://taskbook.nasaprs.com/Publication/index.cfm?action=public_ 123–164 (2012). query_taskbook_content&TASKID=11095 (2017). 33. Cacciapuoti, L. & Salomon, C. Atomic clock ensemble in space. J. Phys. Conf. Ser. 52. Cornell, E. Zero-G Studies of Few-body and Many-body Physics. https://taskbook. 327, 012049 (2011). nasaprs.com/Publication/index.cfm?action=public_query_taskbook_content& 34. Williams, J. R., Chiow, S.-W., Yu, N. & Müller, H. Quantum test of the equivalence TASKID=11096 (2017). principle and space-time aboard the international space station. New J. Phys. 18, 53. Dieckmann, K., Spreeuw, R. J. C., Weidemüller, M. & Walraven, J. T. M. Two- 025018 (2016). dimensional magneto-optical trap as a source of slow atoms. Phys. Rev. A 58, 35. Thompson, R. Science envelope requirements document (SERD) for cold atom 3891 (1998). laboratory. Tech. Rep. Jet Propulsion Laboratory (2013). 54. Salim, E. A., Caliga, S. C., Pfeiffer, J. B. & Anderson, D. Z. High resolution imaging 36. Farkas, D. M., Salim, E. A. & Ramirez-Serrano, J. Production of rubidium and optical control of Bose–Einstein condensates in an atom chip magnetic trap. Bose–Einstein condensates at a 1 Hz rate. arXiv:1403.4641v2 (2014). Appl. Phys. Lett. 102, 084104 (2013). 37. Silber, C. et al. Quantum-degenerate mixture of fermionic lithium and bosonic 55. Gokhroo, V., Rajalakshmi, G., Easwaran, R. K. & Unnikrishnan, C. S. Sub-doppler rubidium gases. Phys. Rev. Lett. 95, 170408 (2005). deep-cooled bosonic and fermionic isotopes of potassium in a compact 2D -3D 38. Wang, P., Xiong, D., Fu, Z. & Zhang, J. Experimental investigation of evaporative MOT set-up. J. Phys. B At. Mol. Opt. Phys. 44, 115307 (2011). 87 40 cooling mixtue of bosonic Rb and fermionic K atoms with microwave and 56. Landini, M. et al. Sub-doppler laser cooling of potassium atoms. Phys. Rev. A 84, radio frequency radiation. Chin. Phys. B 20, 016701 (2011). 043432 (2011). 39. Marzok, C., Deh, B., Courteille, P. & Zimmermann, C. Ultracold thermalization of Li and Rb. Phys. Rev. A 76, 052704 (2007). 40. Haas, M. et al. Species-selective microwave cooling of a mixture of rubidium and Open Access This article is licensed under a Creative Commons caesium atoms. New J. Phys. 9, 147 (2007). Attribution 4.0 International License, which permits use, sharing, 41. Taglieber, M., Voigt, A., Aoki, T., Hänsch, T. & Dieckmann, K. Quantum degenerate adaptation, distribution and reproduction in any medium or format, as long as you give two-species fermi–fermi mixture coexisting with a Bose–Einstein condensate. appropriate credit to the original author(s) and the source, provide a link to the Creative Phys. Rev. Lett. 100, 010401 (2008). Commons license, and indicate if changes were made. The images or other third party 42. Campbell, R. et al. Efficient production of large K Bose–Einstein condensates. material in this article are included in the article’s Creative Commons license, unless Phys. Rev. A 82, 063611 (2010). indicated otherwise in a credit line to the material. If material is not included in the 43. Modugno, G. et al. Bose–Einstein condensation of potassium atoms by sympa- article’s Creative Commons license and your intended use is not permitted by statutory thetic cooling. Science 294, 1320 (2001). regulation or exceeds the permitted use, you will need to obtain permission directly 44. Xiong, D., Wang, P., Chen, H. & Zhang, J. Species-selective microwave cooling of a from the copyright holder. To view a copy of this license, visit http://creativecommons. mixture of rubidium and caesium atoms. Chin. Opt. Lett. 8, 351 (2010). org/licenses/by/4.0/. 45. Modugno, G., Modugno, M., Riboli, F., Roati, G. & Inguscio, M. Two atomic species superfluid. Phys. Rev. Lett. 89, 170406 (2002). © The Author(s) 2018 Published in cooperation with the Biodesign Institute at Arizona State University, with the support of NASA npj Microgravity (2018) 16

Journal

npj Microgravity – Springer Journals

Published: Aug 21, 2018

There are no references for this article.