Atom Interferometry test of short range gravity : recent progress
Transcription
Atom Interferometry test of short range gravity : recent progress
Atom interferometry test of short range gravity : recent progress in the ForCa-G experiment Experiment : Matthias Lopez, Obs Cyrille Solaro, Obs Franck Pereira, Obs Theory : Astrid Lambrecht, LKB Axel Maury, LKB Gabriel Dufour, LKB Marie-Christine Angonin, Obs Peter Wolf, Obs 1 Outline • Introduction Inertial sensors with cold atoms, Why gravity needs testing… • State of the art in our lab Vertical lattice, Wannier Stark ladder, Bloch frequency & local gravimetry. Experimental Setup. • The next step Casimir-Polder potential probing, Mirrors in vacuum, Manipulation of atoms in the surface vicinity… 2 Inertial sensors with cold atoms @ SYRTE, in the IACI group. • • • • • Gyrometer (Remi Geiger) Gradiometer (Franck Pereira & Sebastien Merlet) Gyrometer on chip (Carlos Guerrida) Trapped atomic clock on chip (P. Rosenbusch) MIGA (GW) (Geiger and collaborators @ LP2N, LBB… and more) • Gravimeter (Franck Pereira & Sebastien Merlet) σ𝑔 = 5.7 × 10−9 @ 1𝑠 𝑔 3 Gravitation, why does it need testing ? Two powerful theories : Standard Model : Electromagnetic, weak and strong & General Relativity: Gravitation. These two theories are fundamentally incompatible. Unifying models with higher dimensionality predict that gravitational force should differ at short range. (Adelberg, Ann. Rev. Part. Sci 53, 77, 2003) They predict neither range, nor magnitude… merely constraints. 4 Constraints Klein-Gordon equation 2𝑐2 𝑚 𝜕𝜇𝜕𝜇 + 2 𝑈 = 0 ℏ Yukawa type potential: 𝑈(𝑟) = 𝐶 𝑟 𝑒 − 𝑚𝑐 𝑟 ℏ This formalism is used to parameterize the deviation, it yields no physical content but range λ and amplitude α 𝑈𝑁𝑒𝑤𝑡𝑜𝑛 𝑟 𝐺𝑀𝑚 − = 1 + 𝛼𝑒 λ 𝑟 5 Gravitation, measurements at different scales Long range (103 to 1011 m): Telemetry (satellite or lunar) (Ciufolini, Science 279, 2100 (1998)) Planetary Orbitography (Kolosnitsyn, Gen. Rel. Grav. 36, 1619 (2004)) Pulsars (Will, Astrophysics and Space Science 63, 731 (2004)) Medium range ( ~ meters): Free fall tower (Eckhart, Phys. Rev. Lett., 60, 2567 (1988)) Short range ( < meter): Torsion pendulum (Hoskins, Phys. Rev. D., 32, 3084 (1985)) Optical interferometry (Smullin, Phys. Rev. D 72, 122001 (2005)) Casimir effect (Decca, Phys. Rev. Lett. 78, 5(1997)) 6 Some visual insight on constraints Large Scale Small Scale log10a Lab Satellite log10l (m) LLR Orbitometry E. Fischbach, R. Hellings, & al. (2003) A. Geraci et al., Phys Rev D 78, 022002 (2008) 7 Principle of the experiment, Hamiltonian. “A vertical trapped atomic interferometer close to a surface” g Energy Mirror B λl/2 = 266 nm ll / 2 z Site m Rb Atoms P 2 U lattice H 1 cos 2klattice z ma gz 2ma 2 Kinetic energy Bloch Frequency : Trapping potential Gravity λ𝑙 ℎν𝐵 = 𝑚𝑎𝑔 = ℎ × 568.05 𝐻𝑧 2 8 Principle of the experiment, solutions. Eigenstates : Wannier Stark states ∀𝑚 , |𝜑𝑚 > = Eigenvalues Em, With the following property : 𝐸𝑚 − 𝐸𝑚+∆𝑚 = ∆𝑚 × ℎν𝐵 “Wannier-Stark ladder” Knowledge of νB yields knowledge on the local field (gravitationnal and more…) ! 9 Principle of the experiment, interferometry. Two counter-propagating Raman beams couple : • Internal degrees of freedom : Rb hyperfine structure • External degrees of freedom : position on lattice π/2 νHFS π/2 RamseyTime T t MIRROR MIRROR Δm∙νB MIRROR m m+Δm We then measure populations in both hyperfine states m+Δm g 𝑃𝑒 𝐶 = 1 + cos Δϕ 𝑃𝑒 + 𝑃𝑔 2 where m 𝑈𝑚+Δ𝑚 − 𝑈𝑚 Δϕ = ×𝑇 ℏ m Which yields the Bloch frequency νB ! 10 Current experimental setup 1. Cold atoms in a 3D Magneto Optical Trap 3D 107 atoms in 500ms @ 2μK (Bonus step : Evaporative cooling) 2. 532nm 7W Laser, 800 μm waist vacuum chamber Provides the vertical lattice 3. 1064 nm 500mW Laser, 200 μm waist Provides transverse confinement lattice lattice 4. 2 counter-propagating Raman beams Allows for coherent superposition of wave packets, suitable for interferometry π/2 MOT 3D π/2 MixTrap up to 3 s Detection time k1 k2 MOT 3D k1 Measuring the Bloch Frequency νB Coherent superposition of states on site m and m+Δm Verified with Rabi oscillations. TramseyU==100ms 1.8 Er ↔ 1 fringeΔm every = +610Hz Interferometric fringes within an enveloppe. We locate the central fringe. νCF = Δm ∙ νB Today : 𝜎𝑔 𝜎𝜈𝐵 = = 2.2 10−6 @ 1𝑠 𝑔 𝜈𝐵 Corresponds to 0.1 mHz in 100 seconds Integration time ! 12 Phenomenology in the vicinity of a conducting surface Utot = Ugrav + UCP + UYukawa Gravitationnal Potential g We have an interferometer that measures g locally Casimir, surf-surf ℏ𝑐𝜋2 𝑈𝐶 = −𝐴 240𝐿3 Casimir-Polder Interaction Casimir-Polder, surf-dipole Surface L Atom 𝑈𝐶𝑃 3ℏ𝑐𝛼0 =− 8𝜋𝐿4 L By precisely calculating and measuring those 2 effects : Deviation (the quest…) New constraints on range λ and amplitude α 13 Consequence of CP on energy levels in the vicinity of a mirror. MIRROR, M Energy |e> |g> νB νB+νCP Position 14 Numerical Calculations of the Casimir Polder potential Pelisson,PRA 86, 013614 (2012) Energy Shift due to Casimir-Polder Interaction ΔνCP (x 3.77 kHz) Real C-P Potential “Naïve” C-P Potential ΔE = ~2 Hz ! z atom distance from mirror (in site units) 4 orders of magnitude higher than our resolution @100s ! 15 A quick break ! What we have : A trapped interferometer capable of measuring local potentials, with enough resolution to probe with great accuracy the CP potential. The vertical lattice reflection mirror, which is currently outside the vacuum chamber needs to be placed inside ! What we need : The means to transport atoms close to the surface, in a well controlled manner. 16 Mirror inside (coming soon…) MIRROR At the moment, we have one ultra low pressure vacuum chamber (10-10 mbar) The mirror is outside We will in the next months add another science chamber on top. • 4 mirrors (1/2 inch) on translation stage • Large optical access • Electric field control • Independant vacuum 532 nm The “naked” vacuum cell 17 Moving the atoms from one vacuum cell to the other, the idea. Atom elevator (aka Bloch lift) By controlling the frequency difference between 2 laser beams, we effectively create a moving lattice, accelerating and decelerating the Rb Atoms. Parameters : 40 GHz detuning 100 mW/beam U = 100 Er a = 120 g 18 Moving the atoms from one vacuum cell to the other. Ben Dahan, PRL 76, 4508 (1996) Cadoret, PRL 101, 230801 (2008) Efficiency limited by • Size of beam < size molasse • Temperature of atoms • Spontaneous emission Test on atoms from an optical molasse 19 Is loading the MixTrap from a Magneto-Optical Trap enough ? The Problem: • • • 60000 atoms populate 4000 sites 15 atoms per site, covering a length of 1mm. 2 μK temperature, which implies low efficiency of the Bloch Elevator (3 atoms per site once lifted) The requirements: • • • • Lots of atoms too maintain decent signal at the detection Populate smaller span of sites More atoms per site Reduce temperature “We need to load the MixTrap from a cooler, smaller and denser sample” The solution: 20 Reaching our goal through evaporative cooling f1=300mm 300+150mm f=150mm 100 W 1064nm laser 2 AOM to control beam power 100+150mm f2=100mm 196µm 35.5µm f=150mm -1 order +1 order 300 mm 300 mm AOM f=150mm Vacuum chamber 150mm f=150mm Create a cigar shaped trapping dipolar potential: Width ~ 30 um Length ~ 150 um AOM 150mm 110 mm 172x48.5µm EVAPORATIVE COOLING 21 Benefits of evaporative cooling Within a few seconds, we increase phase-space density by : Lower temperature Better space density Fewer states populated in transverse confinement → Better contrast at longer Ramsey time ↘Better resolution on the Bloch Frequency νB We now have 1011-1012 atoms/cm3 → More atoms per site, less sites are populated ↘ 40000 atoms in σ = 4 sites (1 um) ↘ We can expect better site adressability close to the surface. Loading from Molasse Loading from Dipolar Trap Number of atoms 60000 40000 Sites populated 2000 20 Atoms per site 30 2000 Preparation time 500 ms to 1s 3 seconds 22 The unsuspected benefit of higher densities. What we would expect : Different collisional shifts in different sites should kill the contrast at long times, due to spin dephasing. (νcoll = 0,4Hz for 1012 at/cm3) “Identical Spin Rotation Effect” What we see for Δm = 0 : Deutsch, PRL 105, 020401 (2010) Collision induced spin rephasing : 23 Discerning CP from a possible deviation to Newton’s law Numerically: By properly modelling the Casimir Polder potential induced by the di-electric mirror on the atomic dipole. Main Challenge : Mirror is not a perfect conductor, its complex permittivity needs to be well characterized. A. Lambrecht and collegues (LKB) Calculated CP potential is then substracted from measurement → deviation(?) Experimentally: It’s easy to work with alternatively with 2 Rubidium isotopes : 87Rb & 85Rb They have the same atomic polarizability α0. However their masses differ, m87/m85 ≈ 87/85 Same experiment with 87Rb & 85Rb, then ‘87’ - ’85’ « What we have left ‘should’ behave like gravitational force » We have 4 mirrors slots : we have control over the test masses ! 24 At the end of the day… By inserting a mirror inside and properly controlling our site populations close to this mirror we can conservatively expect : Explore the λ ≈ 10 μm range Where CP < 10-2 Hz Explore the λ ≈ 0,2 - 1 μm range With differential measurements 25 Conclusion and perspectives • We expect our ultra-cold Rb atoms to provide us with a great tool to probe short range forces with great accuracy. • A unique tool to probe Casimir forces • The means to discern a possible 5th force… or at least set new constraints. Short term prospects: • Insert mirror in vacuum • Transport atoms close to the mirror surface • Perform interferometric potential measurement at short ranges 26 Thank you for your attention ! Franck Pereira Cyrille Solaro Peter Wolf Astrid Lambrecht The Atomic Interferometry and Inertial Sensors @ SYRTE, Paris Observatory 27