A passively pumped vacuum package sustaining cold atoms for more than 200 days
Bethany J. Little, Gregory W. Hoth, Justin Christensen, Chuck Walker, Dennis J. De Smet, Grant W. Biedermann, Jongmin Lee, Peter D. D. Schwindt
aa r X i v : . [ phy s i c s . a t o m - ph ] J a n A passively pumped vacuum package sustaining cold atoms for more than 200 days
Bethany J. Little, ∗ Gregory W. Hoth, Justin Christensen, ChuckWalker, Dennis J. De Smet, Jongmin Lee, and Peter D. D. Schwindt
Sandia National Laboratories, Albuquerque, NM 87185, USA
Grant W. Biedermann
Department of Physics and Astronomy, University of Oklahoma, Norman, Oklahoma 73019, USA (Dated: January 11, 2021)Compact cold-atom sensors depend on vacuum technology. One of the major limitations tominiaturizing these sensors are the active pumps—typically ion pumps—required to sustain the lowpressure needed for laser cooling. Although passively pumped chambers have been proposed as asolution to this problem, technical challenges have prevented successful operation at the levels neededfor cold-atom experiments. We present the first demonstration of a vacuum package successfullyindependent of ion pumps for more than a week; our vacuum package is capable of sustaining a cloudof cold atoms in a magneto-optical trap (MOT) for greater than 200 days using only non-evaporablegetters and a rubidium dispenser. Measurements of the MOT lifetime indicate the package maintainsa pressure of better than 2 × − Torr. This result will significantly impact the development ofcompact atomic sensors, including those sensitive to magnetic fields, where the absence of an ionpump will be advantageous.
I. INTRODUCTION
Atomic sensors, which make use of the inherent pre-cision of atomic energy levels, are moving from the lab-oratory, where they have pushed the limits of precisionmeasurements, [1–3] to applications in the field, wherethe requirements are shifted toward portability, reliabil-ity, and integration.[4] Successful miniaturization of vac-uum technology for cold-atom sensors will impact a widerange of applications including gravimeters, accelerome-ters, gyroscopes, clocks, magnetometers, and gravity gra-diometers.These sensors must include an ultra-high vacuum(UHV) system capable of sustaining a cloud of coldatoms. This cloud is often produced using a magneto-optic trap (MOT). With existing UHV technology, boththe pumping apparatus and the gauges require many cu-bic centimeters at best,[5] in addition to the chamber it-self. While passively pumped vacuum packages utilizingnon-evaporable getters (NEGs) have been proposed,[6, 7]there has been doubt as to whether or not they aresustainable over the time scales needed for cold-atomsensors.[5, 8] Getters have been used in miniature ion-clock chambers,[9, 10] however these systems take advan-tage of a getter-pump’s inability to pump noble gases, asevere disadvantage for applications involving MOTs.We present a vacuum chamber that uses passive pump-ing to maintain pressures sufficient for a MOT in excessof 200 days. Using the MOT as a rough pressure gauge,we estimate that the system sustains a background pres-sure of better than 2 × − Torr.[11, 12] Although thispressure is relatively high for cold-atom systems, it issufficient for applications such as high data rate atom ∗ [email protected] copper pinch-off tubegetters dispensers(a) (b)(c) FIG. 1. (a) The vacuum package consists of a titanium bodywith six sapphire windows and other components laser-weldedonto it. (False transparency reveals the inside components ofthe arms, shown in the insets in more detail.) A copper pinch-off tube allows pumping down with standard turbo and ionpumps and has been successfully pinched off to create a cold-welded seal that preserves the vacuum. (b) Two rubidiumdispensers are held in place with alumina spacers. (c) Detailof one of the non-evaporable getters. interferometry.[13] While efforts are being made by oth-ers to miniaturize vacuum systems for the purpose ofatomic sensors,[7, 8, 14] to our knowledge this is the firstdemonstration of a passively pumped chamber sustaininga MOT for more than a week.[15]
II. VACUUM DESIGN AND FABRICATION
The primary challenge of the design is to make a cham-ber that can sustain a high level of vacuum without de-pendence on active pumping, such as an ion pump, whileallowing the optical access needed for cold atom exper-iments. We utilize non-evaporable getters, which pas-sively pump the chamber by means of chemisorption.[16]Since the NEGs do not pump rare gases, any heliumpermeation will limit the lifetime of the vacuum. Al-though it is easy to find materials for the body of thechamber that will not allow helium permeation, findinga transparent material for the windows with this prop-erty is more challenging. Aluminosilicate [7] and alumina[17, 18] perform better than borosilicate, however the lat-ter presents some fabrication challenges. We utilize sap-phire, which has no documented helium permeation thatthe authors are aware of, and use C-cut windows to min-imize birefringence and maximize strength. The body ofthe package is fabricated from commercially pure tita-nium since it matches well with the coefficient of thermalexpansion of sapphire, has a much lower hydrogen out-gassing rate than stainless steel,[19] and is nonmagnetic.The configuration of the vacuum package is shownin Figure 1. The sapphire windows (MPF Prod-ucts, Inc.) braised into titanium frames are laser-welded into the electropolished titanium package body.Four corner tubes support operation: two with get-ters (SAES St172/HI/7.5-7), one with two dispensers(SAES Rb/NF/3.4/12FT10+10), and a copper tubewhich is connected to a standard UHV pumping appa-ratus prior during initial pump-down and testing, but islater pinched off to form a cold-welded seal. The gettersand dispensers are housed in commercially pure titaniumtubing and are connected to the outside of the chambervia custom electrical feedthroughs, as shown in Figure 1.The external volume of the package is approximately 70mL.The entire assembly is baked in a vacuum furnaceat 400 ◦ C for seven days at 1 × − Torr.[20] Duringthis time, the package interior volume is evacuated inde-pendently with total pressure and residual gas pressuremonitoring to assure sufficient exhaust parameters areachieved and that unwanted gases from the getters anddispensers are removed as they are electrically activated.Final post-exhaust pressures of 3 . × − Torr are ob-served on the vacuum package near the turbo pump. He-lium leak tests are performed throughout the fabricationand exhaust process to ensure vacuum integrity. On thefinal package, the ion pump current of 0.7 nA gives anestimated pressure of 2 . × − Torr.
III. TEST SETUP AND PROTOCOL
We utilize a MOT to characterize our vacuum pack-age. Not only does this demonstrate the viability of thechamber for cold-atom experiments, but it serves as an
FIG. 2. The vacuum package is tested using fluorescencefrom atoms in a MOT. (a) Six circularly polarized 780-nmlaser beams (four of which are indicated by the large arrows)locked 8 MHz to the red of the F=2 to F’=3 line in Rbprovide cooling. A second laser tuned to the F=1 to F’=2line is mixed in with the axial cooling beams and serves asa repump. Anti-Helmholtz coils (AHC) are mounted directlyon the chamber. We observe loading curves by switching thecurrent to these coils off and then back on. The cloud ofatoms is imaged on a CCD; load curves (b) are obtained fromthe CCD and photodiode (PD). A 795 nm probe beam is spliton a beamsplitter (BS), with one arm passing through the in-terior of the chamber, while the other serves as a reference.The difference measurement of the absorption probe is sentfrom the balance detector (BD) to a lock-in amplifier; theamplitude of the signal shown in (c) is used to calculate therubidium pressure. indication of the evolution of the background pressureover time.The trap loading dynamics are well described by:[11,12] dNdt = R − Γ N, (1)where N is the number of atoms in the MOT, R is theloading rate, proportional to the rubidium pressure P Rb ,and Γ = γ Rb P Rb + γ bk P bk + Γ is the loss rate. This lossrate depends on the rate γ Rb of collisions with rubidium,the rate γ bk of collisions with other background gasses ofpressure P bk , and a density-dependent factor Γ account-ing for two body collisions within the cold cloud. In atypical MOT, the density of the trapped atoms is limitedby light scattering forces; in this constant density limitΓ can be approximated as constant and one obtains anexponential loading curve: [11, 12] N = N (1 − e − t/τ ) , (2)where N = Rτ is the total number of atoms loadedwith time constant τ = 1 / Γ. N and τ are measuredby observing the fluorescence of the atoms as a MOTis loaded.[21] A representative loading curve and fit isshown in Figure 2 (b), which was collected after switch-ing on the anti-Helmholtz coils. The CCD counts arecalibrated to determine the number of atoms.[21]Arponthip, et al.[11] showed that τ can be used to es-timate the background pressure in typical UHV systems.Specifically, the pressure inside the chamber is well ap-proximated by P vacuum < (2 × − Torr · s) /τ. (3)Despite a few caveats,[11] this gives a good upper boundon the background pressure of the vacuum system. Wealso monitor the pressure of Rb in the vacuum packagedirectly using a probe laser on the D1 line.A schematic of the test setup is shown in Figure 2.The 780 nm cooling laser is locked 8 MHz to the red ofthe F=2 to F’=3 resonance of Rb. The repump light isresonant with the F=1 to F’=2 transition of Rb. Bothlasers are coupled into polarization maintaining fibersand distributed to the vacuum test setup via splitters.A CCD and a photodiode are used to measure loadingcurves. The majority of the data presented utilizes theresults from the CCD; the photodiode serves to confirmparticularly short loading times (Figure 5). The trappingmagnetic field is generated with a pair of circular anti-Helmholtz coils which are switched via software control.This method allows for background subtraction of flo-rescence from atoms in the chamber, compared to thoseloaded into the trap.The density of rubidium in the chamber is mea-sured via the absorption of another laser beam as itis swept through the D1 F=3 to F’=2,3 transition of Rb, for which the cross-section is calculated to be3 . × − m .[22] From this density, a pressure is cal-culated. We use a balanced detector to reduce the noisedue to power fluctuations. The signal to noise is fur-ther improved by use of a lock-in detector; the currentof the laser is modulated at 100 kHz while it is sweptacross the resonance at 0.5 Hz. The resulting amplitudeof the dispersive lock-in signal shown in Figure 2 (c) isproportional to the absorption of the probe.The vacuum package was initially set up with an ionpump. After establishing testing procedures, the ionpump was switched off. Following successful operationwith the pump off for a month, the ion pump wasswitched back on in preparation for pinching off the cop-per tube. During the month of pumpless operation, theloading times and MOT atom number were around 100 (b)(a)(c) (d) stabi lizedthermalizing FIG. 3. Following pinch-off, (a) the number of atoms N and(b) the characteristic loading times τ of a MOT in the pas-sively pumped chamber are monitored over the course of 200days. Other activity on the optical table required the laserenclosure curtains to be opened during the day, causing driftslike the one shown in (d), which shows the loading time overthe course of a long weekend. Many of these transients havebeen removed from (a) and (b) to better show the trends.Major variations due to large changes in the rubidium dis-penser current are highlighted in red, with the correspondingdispenser current used shown in (c). ms and 7 × , respectively; prior to pinch-off, they werearound 1 s and 7 × . A pneumatic pinch-off tool (CPSHY-187-F) was used to pinch off the copper tube. Re-sults of the measured MOT loading parameters followingpinch-off are shown in Figure 3. IV. RESULTS
The primary result of this work is the demonstrationthat a passively-pumped vacuum chamber can supportcold atom physics experiments for months. Figures 3(a) and (b) show the number of atoms in the trap N and the loading time constant τ over a period of 200days following pinch-off. Each data point is obtained byfitting Equation 2 to a measured loading curve, as in 2(b). Experiments involving large changes of the dispenser FIG. 4. The pressure of the background vapor in our vacuumpackage is estimated using Equation 3, while the pressuresof both isotopes of rubidium in the chamber are estimatedusing the absorption probe measurement. Data is shown sincecalibration of this measurement on day 60. current are highlighted in red. The dispenser current isshown in 3(c). Using the absorption probe and Equa-tion 3, we estimate the pressure of rubidium and thebackground gases in our chamber, as shown in Figure 4.After initial changes immediately following the pinch-off,the MOT loading parameters change relatively slowly, in-dicating that the passive pumping is able to maintain avacuum in the system for many months. Variation in theplots is due to a number of factors, which we divide intoshort-term transients and long-term trends.The short day-to-day variability in MOT loading pa-rameters is dominated by temperature changes caused byopening and closing the curtained enclosure around theoptics table and other work going on in the lab. This ishighlighted in the example shown in Figure 3 (d), whichshows a data run taken over a long weekend; there is aninitial thermalization period which we exclude from ouranalysis. We attribute many of the short-term transientsto a combination of temperature-caused misalignmentsand changes in the rubidium pressure due to ambienttemperature. To test the temperature dependance, weplaced a temperature logger on the table next to the testsetup. Temperature changes of a few tenths of a degreeresult in significant changes to both the atom number N and the loading time constant τ . These variations areconsistent with the observed variations in the Rb pres-sure. More details can be found in the SupplementalMaterial [23].Long-term trends are both more difficult to explainand more interesting. The variation in the average val-ues in Figure 3 (a-b) may be a result of hysteresis inthe alignment in various parts of the test setup, as wellas disturbances to the alignment due to other activity inthe lab. The alignment of the system has been optimizedseveral times to maximize the atom number. These re-alignments tend to cause step changes in N, for example near day 140 in Figure 3 (a). Our absorption measure-ment of the rubidium density indicates an increase in theamount of rubidium in the chamber over the 200 days;this likely plays a role in the increased MOT atom num-ber and decreased loading time seen over this period.For example, from day 60 to 180, the rubidium pressureincreased by around 1 . × − Torr (Fig 4); based onan estimated loss coefficient for Rb-Rb collisions,[11] thiscould contribute to the MOT loss rate Γ = 1 /τ by 0.6s − . During the same time, τ decreases from 120 ms to90 ms (Fig 3), corresponding to an increase ∆Γ ≈ − in the loss rate. Thus while the increase in rubidiumplays a role, there are likely other effects contributing tochanges in τ .In order to understand the role of the rubidium dis-penser in our vacuum package, several dispenser currentvariation experiments were performed. The behavior of N and τ during these experiments suggests that the Rbdispenser plays a significant role in maintaining the pres-sure. The result of one of these tests is shown in Fig-ure 5. Naively one would expect from Equation 1 thatwhen the dispenser is turned off, the number of trappedatoms would begin to decrease as the amount of rubid-ium decreases, while the loading time would increase.[12]Surprisingly, the two parameters trend in the same direc-tion; we hypothesize that this is indicative of a dispenserpumping effect—due to the gettering material of the dis-penser, and possibly also due to the increased number ofalkali atoms, which may serve to reduce the pressure ofother gases in the chamber.This can be seen by plotting the background pressureestimated from loading curve fits and Equation 3 along-side the results of the absorption measurement, in Fig-ure 5 (b). When the dispenser is turned off (day 94),the increase in the background pressure appears to bea larger effect than the decrease in rubidium pressure.When the dispenser is turned back on, there is a spikein background pressure: this effect has been noted byothers, and is likely due to a release of gas from boththe chamber walls and the dispenser material as the dis-penser temperature increases.[24] After some time, thepressures return to their values prior to the cycling ex-periment.While the dispenser pumping effects are significant, itis worth noting that this experiment also demonstratesthat even after three days of leaving the dispenser off,the chamber still supports a cloud of around 5 × coldatoms. V. CONCLUSION
We have demonstrated the successful design and test-ing of a portable vacuum package that can sustain vac-uum levels low enough for cold atoms using only passivepumping via non-evaporable getters. The success of ourdesign is undoubtedly due to a combination of factors,including the low helium-permeability of the materials
FIG. 5. The rubidium dispenser is cycled off and on betweendays 93 and 98 at times marked with vertical lines. (a) Asexpected, the number of atoms in the MOT drops, althoughthis number trends toward a non-zero cloud size, indicatingthat the chamber could sustain operation without the dis-pensers for a significant period of time. The loading time alsodecreases; this is used to estimate the rising background pres-sure (Eq. 3) shown in (b), along with the rubidium pressurescalculated from the absorption measurement. and the fabrication procedures. We are in the processof building and testing other vacuum chambers with dif-ferent parameters; the results of these future tests willshed light on which parts of the design and fabricationare most critical.Our design has a number of advantages for atomic sen-sors. While some have proposed the use of other non-magnetic ion pumping mechanisms,[8] eliminating theneed for an ion pump altogether presents a clear advan-tage. We have successfully driven 6.8 GHz microwavetransitions in rubidium in the chamber, demonstratingthat the metal body is not an obstacle to such atomicstate manipulation. Finally, in contrast to chip-focused designs,[14] the six windows allow optical access for thecounter-propagating beams typically used in atom inter-ferometer applications.The behavior of the MOT loading time and atom num-ber in response to different changes, such as the currentin the dispenser or the temperature of the environmentwill be the subject of future study. In such a small cham-ber, subtleties arising from the likely pumping of the dis-penser present both challenges and advantages that wewould like to understand better. The dispenser requires2.15 A current to maintain optimal conditions, or 2.22 Wof power. Continued investigation is required to developtechniques to achieve substantially reduced power con-sumption while maintaining the appropriate alkali[24, 25]and background pressures.Our result represents significant progress toward re-ducing the size, weight, and power consumption of atomicsensors. We expect that it will drive the development ofmore robust sensors which can be used in a wide range ofapplications, from fundamental physics such as measure-ments of gravity in space, to more application-focuseddevelopments in civil engineering. It is particularly well-suited for inertial sensing devices, where the demand ishigh for a robust and compact vacuum package.
ACKNOWLEDGMENTS
The authors would like to Melissa Revelle for her con-tributions to the testing, and members of John Kitching’sgroup at NIST for their ideas and feedback. This workwas supported by the Laboratory Directed Research andDevelopment program at Sandia National Laboratories.Sandia National Laboratories is a multimission labora-tory managed and operated by National Technology &Engineering Solutions of Sandia, LLC, a wholly ownedsubsidiary of Honeywell International Inc., for the U.S.Department of Energy’s National Nuclear Security Ad-ministration under contract DE-NA0003525. This paperdescribes objective technical results and analysis. Anysubjective views or opinions that might be expressed inthe paper do not necessarily represent the views of theU.S. Department of Energy or the United States Govern-ment.
AUTHOR’S CONTRIBUTIONS
G. W. Hoth and B. J. Little contributed equally to thiswork. [1] R. H. Parker, C. Yu, W. Zhong, B. Estey, and H. M¨uller,Measurement of the fine-structure constant as a test ofthe standard model, Science , 191 (2018). [2] G. Rosi, F. Sorrentino, L. Cacciapuoti, M. Prevedelli,and G. Tino, Precision measurement of the newtoniangravitational constant using cold atoms, Nature , 518(2014). [3] S. Brewer, J.-S. Chen, A. Hankin, E. Clements, C.-w.Chou, D. Wineland, D. Hume, and D. Leibrandt, Al+ 27quantum-logic clock with a systematic uncertainty below10- 18, Physical review letters , 033201 (2019).[4] K. Bongs, M. Holynski, J. Vovrosh, P. Bouyer, G. Con-don, E. Rasel, C. Schubert, W. P. Schleich, andA. Roura, Taking atom interferometric quantum sen-sors from the laboratory to real-world applications,Nature Reviews Physics , 731 (2019).[5] A. Basu and L. F. Vel´asquez-Garc´ıa, An electrostaticion pump with nanostructured si field emission electronsource and ti particle collectors for supporting an ultra-high vacuum in miniaturized atom interferometry sys-tems, Journal of Micromechanics and Microengineering , 124003 (2016).[6] J. A. Rushton, M. Aldous, and M. D.Himsworth, Contributed review: The feasi-bility of a fully miniaturized magneto-opticaltrap for portable ultracold quantum technology,Review of Scientific Instruments , 121501 (2014).[7] A. T. Dellis, V. Shah, E. A. Donley, S. Knappe, andJ. Kitching, Low helium permeation cells for atomic mi-crosystems technology, Optics Letters , 2775 (2016).[8] J. Sebby-Strabley, C. Fertig, R. Compton, K. Salit,K. Nelson, T. Stark, C. Langness, and R. Livingston, De-sign innovations towards miniaturized gps-quality clocks,in (IEEE, 2016) pp. 1–6.[9] G. K. Gulati, S. Chung, T. Le, J. Prestage, L. Yi,R. Tjoelker, N. Nyu, and C. Holland, Miniatured and lowpower mercury microwave ion clock, in (IEEE,2018) pp. 1–2.[10] Y.-Y. Jau, H. Partner, P. Schwindt, J. Prestage, J. Kel-logg, and N. Yu, Low-power, miniature 171yb ion clockusing an ultra-small vacuum package, Applied PhysicsLetters , 253518 (2012).[11] T. Arpornthip, Vacuum-pressure measurement us-ing a magneto-optical trap, Physical Review A ,10.1103/PhysRevA.85.033420 (2012).[12] R. W. G. Moore, L. A. Lee, E. A. Findlay,L. Torralbo-Campo, G. D. Bruce, and D. Cas-settari, Measurement of vacuum pressure witha magneto-optical trap: A pressure-rise method,Review of Scientific Instruments , 093108 (2015).[13] A. V. Rakholia, Dual-axis high-data-rate atom interfer-ometer via cold ensemble exchange, Physical Review Ap- plied , 10.1103/PhysRevApplied.2.054012 (2014).[14] J. McGilligan, K. Moore, A. Dellis, G. Martinez,E. de Clercq, P. Griffin, A. Arnold, E. Riis, R. Boudot,and J. Kitching, Laser cooling in a chip-scale platform,Applied Physics Letters , 054001 (2020).[15] R. Boudot, J. P. McGilligan, K. R. Moore, V. Maurice,G. D. Martinez, A. Hansen, E. de Clercq, and J. Kitch-ing, Enhanced observation time of magneto-optical trapsusing micro-machined non-evaporable getter pumps, Sci-entific Reports , 1 (2020).[16] St 171 ® and St 172 - Sintered Porous Getters , SAES Get-ters (2020).[17] J. F. O’Hanlon, A user’s guide to vacuum technology (John Wiley & Sons, 2005).[18] W. Perkins, Permeation and outgassing of vacuum mate-rials, Journal of vacuum science and technology , 543(1973).[19] M. Takeda, H. Kurisu, S. Yamamoto, H. Nakagawa, andK. Ishizawa, Hydrogen outgassing mechanism in titaniummaterials, Applied surface science , 1405 (2011).[20] We are uncertain that baking the miniature vacuumwithin a larger vacuum furnace is necessary. We did thisto ensure there was not helium permeated into any of thematerials, but none of the materials should allow any he-lium permeation so this step may have been unnecessary.[21] D. A. Steck, Rubidium 85 D line data, available onlineat http://steck.us/alkalidata (2019).[22] P. Siddons, C. S. Adams, C. Ge, and I. G. Hughes, Abso-lute absorption on rubidium D lines: comparison betweentheory and experiment, Journal of Physics B: Atomic,Molecular and Optical Physics , 155004 (2008).[23] See supplemental material at [url will be inserted by pub-lisher] for further data and discussion on the role of tem-perature variance in our measurements. All files relatedto a published paper are stored as a single deposit andassigned a Supplemental Material url. This url appearsin the article’s reference list.[24] R. N. Kohn Jr, M. S. Bigelow, M. Spanjers, B. K. Stuhl,B. L. Kasch, S. E. Olson, E. A. Imhof, D. A. Hostut-ler, and M. B. Squires, Clean, robust alkali sources byintercalation within highly oriented pyrolytic graphite,Review of Scientific Instruments , 035108 (2020).[25] S. Kang, R. P. Mott, K. A. Gilmore, L. D. Sorenson,M. T. Rakher, E. A. Donley, J. Kitching, and C. S.Roper, A low-power reversible alkali atom source, Ap-plied Physics Letters110