Cold Magnesium Atoms for an Optical Clock Tanja Mehlstäubler Jan Friebe Volker Michels Karsten Moldenhauer Nils Rehbein Dr. Hardo Stöhr Dr. Ernst-Maria Rasel Prof. Dr. Wolfgang Ertmer Institute of Quantum Optics University of Hannover Welfengarten 1 D-30159 Hannover Germany
The Mg - Atom Interferometry group in Hannover Jan Friebe Volker Michels Tanja Mehlstäubler Karsten Moldenhauer Nils Rehbein former members: Holger Wolf Dr. Jochen Keupp Dr. Albane Douillet Dr. Hardo Stöhr Dr. Ernst-Maria Rasel Prof. Wolfgang Ertmer
Outline Brief history of frequency standards Ramsey-Bordé interferometery in Mg How to get to µk temperatures? (A) 25 Mg and heating in MOT (B) Quench cooling (C) 2-photon resonances Summary & Outlook
A brief history of frequency standards microwave standards 1950 : Ramsey develops seperated field method Ø atomic beam clocks surpass quarz 1989 : a Mg atomic beam frequency standard Bava et al., Appl. Phys. B, 48, p.495, (1989) 1989 : first fountain (Na): ν = 2 Hz Kasevich et al., PRL, 63, p.612, (1989) optical standards neutral alkaline-earth (like) atoms Æ high S/N 1992 : the Mg Ramsey interferometer Sterr et al., Appl. Phys. B, 54, p.341, (1992) 1999 : narrow line cooling in Sr : ρ ~ 10-2 in MOT Katori et al., PRL, 82, p.1116, (1999) 2001 : quench cooling in Ca: T MOT ~ 6 µk Binnewies et al., PRL, 87, 123002, (2001) Curtis et al., Phys. Rev. A, 64, 031403, (2001) 2003 : Yb BEC via narrow line cooling! Takasu et al., PRL, 91, 040404, (2003) δν/ν 0 = 7µ10-16 σ y = 1.6µ10-14 in 1s (Cs/Rb fountains) frequency σ ν 0 2 1 y ( τ ) = frequency δν/ν 0 = 1.2µ10-14 (Ca) σ y = 7µ10-15 in 1s (Ca/Hg + comparison) ν 0 π Q Stability S N time τ T Accuracy δν/ν 0 time 1/ 2, Q = ν ν
The magnesium atom 24 Mg 457 nm clock transition atomic quality factor Q=2â10 13 potential for T < µk (1) 285 nm cooling transition strong light forces high T D ~ 2 mk MOT 285 nm 80 MHz J=0 1 S 0 1 P 1 1 D 2 3 S 1 3 P 012 3 D 123 4 MHz 881 nm Singlet 462 nm 457 nm, 31Hz - Interferometry Triplet 1 2 0 (1) H. Wallis and W. Ertmer, J.Opt.Soc.Am. B 6, 2211 (1989)
Ramsey-Bordé Interferometry Doppler free Ramsey spectroscopy Ch. Bordé, C.R. Acad. Sci. Ser. B, 284, p.101 (1977) 1 P 1 285 nm MOT 24 Mg 457 nm 31 Hz 1 S 0 3 P 1 laser beam/puls π/2 16 cm gú T D T D eú ULE- resonator intermediate mass spring mountis Eddie current damping P ex cos(2t D ( +δ rec )) 140 mw at 457 nm with dye laser stabilized to high finess cavities 12,5 cm
Ramsey-Bordé Interferometry 1240000 1220000 290 Hz resolution decay of signal count rate [s -1 ] 1200000 1180000 1160000 1140000 1120000 1100000 N = (8700±1000)s -1 S = 16000 Q = 2.26 10 12 σ y (τ= 1s) = 7.64 10-14 T = 860,8 µs r60b 39600 40000 40400 40800 41200 detuning [Hz] 21.05.02 resolution 280 Hz potential stability σ y = 8 x10-14 laser line width υ Laser (170 15) Hz T MOT = 3.8 mk ï atomic motion limits stability ï only 8 % of atoms are excited (τ Pulse = 4 µs)
Temperatures in MOT new set up to study cooling techniques optical access for UV, quenching, interferometry, dipole trapping up to10 8 atoms Mg beam MOT coils (grad B = 130 Gauss/cm) slowing beam Temperature [mk] 8 7 6 5 4 3 2 change of oven temperature misaligned MOT change of laser intensity Mg25 0 2 4 6 8 10 12 14 16 18 20 N 1/3 n 2/3 [10 8 /cm 2 ] Temperature [mk] 16 14 12 10 8 6 4 2 0,0 0,2 0,4 0,6 0,8 1,0 1,2 S total Doppler limit 24 Mg 25 Mg
Sub-Doppler forces in 25 Mg Sub-Doppler cooling of 25 Mg? (Coop. J.Dunn, J.Ye, NIST) 87 Sr Ω/Γ=1 [hkγ] 25 Mg F=3/2 F=5/2 F=7/2 =19 MHz =27 MHz [Γ/k] [Γ/k] F=5/2 γ=80 MHz I sat =450 mw/cm 2 > 10 5 atoms in MOT 25 Mg Ω/Γ=0.25 B=16 MHz γ] γ] [hkγ] γ] [hkγ] Hyperfine Quenching of clock transition 1 S 0 3 P 0 : 90 µhz 0.44 mhz (Porsev u. Derevianko, physics/0312006, Dez.2003) B=0 MHz [Γ/k] [Γ/k]
Sub-Doppler forces in 25 Mg Sub-Doppler cooling of 25 Mg? (Coop. J.Dunn, J.Ye, NIST) 87 Sr Ω/Γ=1 [hkγ] 25 Mg F=3/2 F=5/2 F=7/2 =19 MHz =27 MHz [Γ/k] [Γ/k] F=5/2 γ=80 MHz I sat =450 mw/cm 2 > 10 5 atoms in MOT 87 Sr 25 Mg (s = 2) v c ~ 0.07 m/s v c ~ 0.22 m/s (s=0.13) v c ~ 0.02 m/s v rec = 0.01 m/s v rec = 0.06 m/s Hyperfine Quenching of clock transition 1 S 0 3 P 0 : 90 µhz 0.44 mhz (Porsev u. Derevianko, physics/0312006, Dez.2003) in MOT: v rms = 0.4 m/s v rms = 1.1 m/s
Quench cooling in 24 Mg 4 1 S 0 γ = 4 MHz quench laser 462 nm MOT 285 nm 80 MHz 3 1 S 0 1 P 1 Clock transition 457 nm 31Hz 3 P 1 cooling on narrow lines 3 1 S 0 3 3 P 1 : T rec = 3.8 µk >> T Dopp mg<f light γ min ~90 Hz quench transition 3 3 P 1 4 1 S 0 200 mw @ 462 nm with Ti:Sapph doubled with PPKTP 24 Mg Γ 3 4 1 3 Ω 23, Γ 2 31Hz 2 Ω 12, Γ 1 }γ eff T.E. Mehlstäubler et al., J.O.B 5, p.183 (2003) ρ 22Γeff = ρ22γ1 + ρ33γ3 for s 3 <<1: Γ eff = Γ 1 Ω + Γ 2 23 3 ρ 33 = ρ 22 Ω Γ Γ Γ 2 3 2 23 2 3 ILaser
Search for the quenching transition retro reflector detection (PMT) Mg oven T ~ 410 C Mg beam interaction zones l/2 462 nm excitation 4 1 S 0 l/2 3 3 P 1 457 nm 3 1 S 0
Quench cooling in 24 Mg Kuruzc: Γ 2 = 3.21 10³ s -1 new ab initio calculations: Pal chikov, Derevianko, Fischer (3) Γ 2 = 2.0 x 10 2 s -1 our measurement: Γ 2 ~ 1 x 10 2 s -1 Transfer Efficiency [%] 60 60 50 50 40 30 20 10 0 (larger (larger power power @@ 457nm) 457nm) 0 5200 10400 15 60020 80025 1000 30 P Quench [mw] per beam σ + σ - J=0 J=0 462nm σ - J=1 σ + 457nm theo. transfer efficiency ~ 1% x 10 4 atoms expected @ T=10µK 8% of 10 5 atoms, 70% of x 10 4 atoms lower duty cycle of interferometry (3) private communications, to be published
2-Photon resonances in 24 Mg 3 1 D 2 2.2 MHz 881 nm exact solution of the Bloch equations for a 3 level system ρ 3 (IR) loss 3 P 1,2 3 1 P 1 3 1 S 0 80 MHz UV-MOT ρ 2 (UV) 2-photon process more efficient than 1 photon? W.C. Magno et al., Phys. Rev. A 67, 043407 (2003)
2-Photon resonances in 24 Mg 3 1 D 2 loss 3 P 1,2 3 1 P 1 2.2 MHz 881 nm 80 MHz UV-MOT UV counts e6/s 8 7 6 5 4 3 Resonance in MOT 80 MHz P IR = 1.3 mw σ = 16 MHz 3 1 S 0 2 1 P IR = 44 mw σ = 26 MHz -70-60 -50-40 -30-20 -10 0 10 20 30 40 50 detuning of 881 nm laser [MHz] 2-photon process more efficient than 1 photon? W.C. Magno et al., Phys. Rev. A 67, 043407 (2003)
2-Photon resonances in 24Mg resonances in 1D molasses
2-Photon resonances in 24Mg resonances in 1D molasses
2-Photon resonances in 24 Mg 24 Mg Temperature across 2-photon resonance 14 26 mk 21 mk scan7b 25 12 20 T [mk] 10 8 6 4 T [mk] 15 10 5 2 60 80 100 120 140 ν (IR) in MHz 0 30 45 60 75 90 105 120 ν (IR) in MHz
2-Photon resonances in 24 Mg Populations in excited states ρ 33 UV IR UV fi force at UV transition: 40000 20000 ρ 22 UV -3-4 -4-6 -2-2 -3-2 -2-2 -4-2 -2-2 -1-1 -1-2 -1 1 2 1 1 2 2 4 2 2 2 6 43 43 ρ 22-20000 -40000 v[m/s] positive feedback on faster atoms (bunching) higher friction for low velocities
2-Photon resonances in 24 Mg force at UV transition fi dv = a(v) dt 20000 25000 20000 20000 30000 30000 20000 15000 15000 20000 20000 15000 20000 10000 10000 10000 10000 5000 5000 10000 5000-3 -2-1 5000 1 2 3-3-4-4-2-2 -2-2-1 1 2 1 2 4 23-2 1-2 -2-3 -2-1 -1-10000 1-1 -2-1 -5000 1 1 2 1 2 2-10000 2-5000 -5000-20000 -10000-5000 0.03 0.03 0.03 0.025 0.025 0.025 0.02 0.02 0.02 0.015 0.015 0.015 0.01 0.01 0.01 0.005 0.005 0.005-0.002-0.001-0.001 0.001 0.002 0.002 0.003 0.003 0.004 e.g. for balanced molasses 20000 10000-4 -2 2 4-10000 -20000 0.03 0.025 0.02 0.015 0.01 0.005-0.004-0.002 0.002 integrating eqs. of motion TOF after 100µs molasses T < 100 µk?
Summary & Outlook Mg interferometry limited by temperature of atoms Sub-Doppler cooling forces in 25 Mg are not sufficient Quenching transition @ 462 nm measured Ø load 1% of atoms into QMOT at 9 µk? 2-photon resonances observed Ø Sub-Doppler temperatures possible in 3D-molasses? Ø Channel to populate meta-stable states! Combination with dipole trap?
Summary & Outlook w 0 = 20 µm Τ trap 0.3 mk U dip [MHz] ac-stark-shift Γ Raleigh 0.1 Hz 100 75 50 25 1 S 0 m = 0 π 3 P 1 m = 0 σ +/ m = +1 / -1 σ /+ m = -1 / +1 σ +/ 400 600 800 1000 λ trap [nm] - Yb:YAG 25 W @ 1030 nm - quench cooling into dipole trap - study of collisions -25-50 -75 - trap potentials only identical at crossing Æ λ = 446 nm -100
The thin disc laser Nd:YVO 4 crystal parabolic pump mirror heat sink pump fiber etalons output coupler 700 mw @ 914 nm 40 W @ 808 nm Laser active medium : thin disk (~320 µm) multiple pump light passes through the laser active medium single frequency with 2 etalons : d = 0,6 mm und 4 mm
Frequency doubling 400 mw @ 914 nm w 2 Det. λ / 4 PPKTP crystal 160 mw @ 457 nm w 1 Cavity: Hänsch-Couillaud lock w 1 ~49 µm w 2 ~147µm cavity length = 34 cm incoupling efficiency: 85 % conversion efficieny: 47 %