View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by CU Scholar Institutional Repository University of Colorado, Boulder CU Scholar Physics Faculty Contributions Physics Spring 5-1-2003 Cooling and Trapping of Atomic Strontium Xinye Xu University of Colorado Boulder, [email protected] Thomas H. Loftus University of Colorado Boulder John L. Hall University of Colorado Boulder, [email protected] Alan Gallagher University of Colorado Boulder Jun Ye University of Colorado Boulder, [email protected] Follow this and additional works at: http://scholar.colorado.edu/phys_facpapers Part of the Physics Commons Recommended Citation Xu, Xinye; Loftus, Thomas H.; Hall, John L.; Gallagher, Alan; and Ye, Jun, "Cooling and Trapping of Atomic Strontium" (2003). Physics Faculty Contributions. 63. http://scholar.colorado.edu/phys_facpapers/63 This Article is brought to you for free and open access by Physics at CU Scholar. It has been accepted for inclusion in Physics Faculty Contributions by an authorized administrator of CU Scholar. For more information, please contact [email protected]. 968 J. Opt. Soc. Am. B/Vol. 20, No. 5/May 2003 Xu et al. Cooling and trapping of atomic strontium Xinye Xu, Thomas H. Loftus, John L. Hall, Alan Gallagher, and Jun Ye JILA, National Institute of Standards and Technology and University of Colorado, Boulder, Colorado 80309-0440 Received August 6, 2002; revised manuscript received December 5, 2002 We present a detailed investigation of strontium magneto-optical trap (MOT) dynamics. Relevant physical quantities in the trap, such as temperature, atom number and density, and loss channels and lifetime, are explored with respect to various trap parameters. By studying the oscillatory response of a two-level 1 1 88 S0 – P1 Sr MOT, we firmly establish the laser cooling dynamics predicted by Doppler theory. Measure- ments of the MOT temperature, however, deviate severely from Doppler theory predictions, implying signifi- cant additional heating mechanisms. To explore the feasibility of attaining quantum degenerate alkaline- earth samples via evaporative cooling, we also present the first experimental demonstration of magnetically trapped metastable 88Sr. Furthermore, motivated by the goal of establishing the fermionic isotope 87Sr as one of the highest-quality, neutral-atom-based optical frequency standards, we present a preliminary study of sub- Doppler cooling in a 87Sr MOT. A dual-isotope (87Sr and 88Sr) MOT is also demonstrated. © 2003 Optical Society of America OCIS codes: 140.3320, 020.7010, 270.4180, 300.1030, 300.2530, 300.6210. 1. INTRODUCTION this system ideal for a future highly accurate optical- frequency standard and presents a preliminary study of Alkaline-earth metal atoms possess versatile internal sub-Doppler cooling in a 87Sr MOT. level structure and a diversity of naturally abundant iso- The desire to explore quantum degenerate gas-phase topes, making these systems extremely useful for funda- bosonic and fermionic mixtures has spawned increasing mental studies of Doppler and recoil-limited laser interest in dual-isotope and heteronuclear MOTs.33–36 In cooling,1–20 precision optical-frequency metrology,21–25 the case of atomic Sr, the high natural abundance of both and ultracold collisions.26–28 Specifically, the zero- bosonic (88Sr, 82% abundance) and fermionic (87Sr, 7% nuclear-spin isotopes are free from hyperfine structure abundance) isotopes, coupled with the ability to cool indi- and are thus ideal platforms for direct tests of Doppler vidual isotopes with single-frequency lasers operating on cooling theory20 and quantitative studies of ultracold the 1S – 1P transition, greatly simplifies the process of light-assisted collisions. In addition, the narrow inter- 0 1 generating mixtures of bosons and fermions in a MOT. combination transitions between singlet and triplet spin Moreover dual-isotope Sr MOTs offer the possibility of manifolds permit ultralow-temperature laser cooling12–18 sympathetically cooling relatively hot 88Sr atoms with in the interesting regime where the photon recoil energy sub-Doppler-cooled 87Sr atoms. In Section 4 we describe exceeds the associated Doppler shift. Highly stable the creation of a dual-isotope MOT for fermionic 87Sr and optical-frequency standards have also been demonstrated bosonic 88Sr. with the alkaline-earth intercombination lines.23,25 On As a first step toward studies of photon-free cold colli- the other hand, the nonzero-nuclear-spin isotopes, such sions and radio-frequency evaporation of Sr, Section 5 de- as 87Sr, provide more attractive systems for optical fre- scribes our efforts to trap 88Sr magnetically in the 3P quency standards: The hyperfine structure of the 2 metastable excited state. Currently we find that, al- 1S – 1P transition may permit single-step sub-Doppler 0 1 though the trap population, loading rate, and density cooling to reach possibly tens of ␮K temperatures, while scale are as expected, blackbody radiation associated with the 1S 3P intercombination line will provide a narrow 0 – 0 the present vapor-cell trap limits the trap lifetime to ϳ40 clock transition that is highly insensitive to external 29 3 ms. A beam-loaded apparatus operated at room tem- perturbations. Moreover, the metastable P2 states 30–32 perature will extend the magnetic trap lifetime beyond present in both isotopes permit magnetic trapping several seconds. and potentially evaporative cooling to Bose or Fermi de- generacy. Of the alkaline earths, atomic Sr is one of the most promising systems for these experiments. 88 This paper is organized as follows. In Section 2 we 2. DOPPLER COOLING SR IN A summarize our recent detailed study of oscillation dy- MAGNETO-OPTICAL TRAP namics in a 88Sr magneto-optical trap (MOT). Here we Figure 1 shows a simplified energy-level diagram for 88Sr. 1 1 ␭ ϭ find that although the trap spring constant and damping The S0 – P1 transition ͓ 461 nm, transition line- coefficient are well modeled by Doppler theory, the sample width (FWHM) ϭ 32 MHz] is used to cool and trap 88Sr temperature greatly exceeds Doppler predictions.20 In in a standard six-beam MOT. In this blue 88Sr MOT, at- addition we report optimum values of trapping-beam in- oms can leak out of the trap by radiatively branching 1 3 tensity and detuning that maximize trap density. Sec- from the P1 state to the metastable P2 state via the 87 1 ! 1 ! 3 tion 3 describes the unique attributes of Sr that make P1 D2 P2 decay path. Atoms decaying to the 0740-3224/2003/050968-09$15.00 © 2003 Optical Society of America Xu et al. Vol. 20, No. 5/May 2003/J. Opt. Soc. Am. B 969 1 1 A simplified diagram of the Sr-vapor-cell, S0 – P1 MOT is shown in Fig. 2. Various experimental results re- ported in this article were obtained with appropriate modifications to this basic setup. The 461-nm cooling and trapping light is generated by frequency-doubling the output from a Ti:sapphire laser. The cooling laser detun- ing is controlled by a double-pass acousto-optic modulator (AOM1). The intensity is stabilized by a second AOM (AOM2). A third AOM (AOM3) provides an on-resonance probe beam for absorption measurements. The two re- pumping beams are produced by external-cavity laser di- odes (ECLD), each locked to a stabilized reference cavity. Since the cooling transition linewidths of the alkaline- earth atoms are generally larger than those of the alkali- metal atoms, the magnetic quadrupole field gradient needs to be larger than the corresponding fields in alkali- Fig. 1. Simplified energy level diagram for 88Sr showing the metal MOTs. For the experiments reported here, we ץ main cooling transition at 461 nm, radiative decay channels for typically employ an axial field gradient ( Bz) of 60 G/cm. 1 the P1 excited state, and the relevant repumping scheme. In the absence of repumping lasers, the Sr-MOT loss Numbers in parentheses give the transition Einstein A coeffi- 37,38 ␻ ⌬ 1 1 rate RL is approximated by cients. L ( )isthe S0 – P1 MOT cooling-laser frequency (de- tuning). It/2Is R ϭ R ϩ A 1 !1 B 1 !3 , L 0 ϩ ϩ ⌬ ⌫ 2 P1 D2 D2 P2 1 It /Is 4͑ / ͒ (1) where R0 is the loss rate caused by background gas colli- sions and the second term on the right-hand side of Eq. (1) gives the loss rate that is due to radiative decay from 1 3 the P1 state to the P2 state. It is the total intensity of ⌬ the six trapping beams, and represents detuning. Is ϭ 2 1 43 mW/cm is the saturation intensity of the S0 ! 1 ⍀2 ϭ ⌫2 ⍀ P1 transition (such that /2, being the Rabi frequency), and ⌫/2␲ ϭ 32 MHz is the transition line- 1 width. For Sr the radiative decay rate from the P1 state 1 3 39 to the D state is A 1 ! 1 ϭ 3.85(1.47) ϫ 10 /s, and 2 P1 D2 1 the predicted branching ratio from the D2 state to the 3 40 P state is B 1 ! 3 ϭ 0.33. Figure 3 shows the 2 D2 P2 88 measured Sr-MOT loss rate as a function of It at a fixed 1 1 ⌬ ␦ ϭ ⌬ ␲ ϭϪ Fig. 2. S0 – P1 vapor-cell Sr MOT. AOM, acousto-optic modu- . Here /2 40 MHz, the axial magnetic field ϭ ץ lator; CCD, charge-coupled-device camera; PMT, photomultiplier gradient Bz 53 G/cm, and the Sr vapor pressure pSr is tube; PD, photodiode; ECLD, external-cavity laser diode; PBS, ϫ Ϫ8 3.5 10 Torr (obtained at a cell temperature Tc polarizing beam splitter. Experimental results reported in this ϭ paper are obtained with appropriate modifications to this basic 360 °C). To determine the loss rate, trap filling is ob- setup.