Atoms near surfaces: Heidelberg node Della Pietra Leonardo Physikalisches Institut der Universität Heidelberg Philosophenweg 12, Heidelberg, Germany Siena

Topics Atom chip Loading & cooling Experiments in HD To do

Atom chip: the technology

Features Optical fields –To load and cool atoms chip surface:evaporated gold layer; highly reflective Magnetic traps –External uniform field B bias plus chip wires define traps tight confinement, high integration, mimimal # of external magnets Electric potentials –Additional degree of freedom to manipulate atom potentials independent of magnetic fields address all atoms, independently of spin state Integration with other techniques –Micro magnets tighter confinement, low noise –Nano-optics Single atom detectors (Cavities), optical guides (push beams,…), optical lattices –Analog & digital standard integrated electronics interface ‘quantum hopes’ with ‘classical realities’ –Micro-mechanics, nano-robots

Wire traps + = wire with current I W homogenous bias field B b B- field: guide atoms in states with µ  B(r ) are confined around a field minimum Approximations : m F is constant of motion pointlike atoms with magnetic moment Frisch and Segre, Z.Phys. 75, 610 (1933) Denschlag et al APB 69, 291 (1999) Denschlag et al. APB 69,291 (1999) Cassettari et al PRL 85,5483 (2000) V(r) = -µB(r) = m F µ B g F |B(r)|

Our chips Size: 2.5 x 3 cm 2, up to 5 µm gold layer Chip = mirror; wires defined by grooves Structures down to 1 µm (limit ~100 nm) Current densities > 10 7 A/cm 2 High V > 300V Trap frequencies up to > 1MHz Ground state size to ~10 nm Incorporation of optical elements Chip fabrication: Heidelberg / Weizmann coorporation Grain size: ~100nm 2 mm Chip design detail (light microscope) 100  m Coating Gold structure Isolation layer (SiO 2 ) Silicon Wafer Gold structure up to 5  m thick

The chip mounting UHV mounting: < mbar easy electrical access 2 layers structure: Macor holder with Cu wire structure high currents ( 60A ) bigger trapping volume 3*10 8 atoms 10 to 2 mm from chip Chip reflective surface for MOT µm sized grooves to define wires

Single chamber Cons Pressure dispenser close to chip; stop it to avoid UHV & lifetimes degradation Concept proved by: München / Tübingen / Sussex/ICL / Heidelberg 45 o MOT beams Pros Simpler only 1 chamber no need for push beam Compact More accessible

Optical access Easy optical access Tight Physical chamber wide chip viewing angles 7+1 windows MOT + cooling Imaging (2 low res + 2 high res) Photodiode...

The imaging system Low res cameras Help in tuning MOT & molasses Fringes Fixed camera window, sharp edges Vibrations -> fix camera, lenses, optical fibers Vertical view new chip along 2D chip structures Schneider et al, PRA 67,023612(2003) High resolution cameras Better definition few µm per pixel Horizontal view to estimate distance from chip diffraction from sharp edges bondings: to be moved

Loading & cooling

atom chi p Rb U-MOT 1.Cu-U MOT 3* 10 8 atoms in 40’’ 2.Molasses 10 ms 100  K 3.Optical pumping | F=2, m F =2  1.Cu-Z magnetic trap ~10 8 atoms Chip-Z mag. trap ~ 5*10 7 atoms Loading sequence 1.Compression 2.RF cooling 20 MHz --  100 kHz 0.8 *10 5 atoms in BEC BEC + thermal atoms

Z wire gives tight & asymmetric magnetic trap Reduced transfer efficiency Reduced heating Z slab opens up & symmetrizes the trap But higher current needed Go for Z-slab? Wires Vs slabs U wire gives poor quadrupole trap U slab increases trapping volume Becker, Diploma Thesis, 2002

Lifetimes Single MOT setup & Cu-Z Dispenser close to chip shortens Cu-Z lifetime: Leave 5‘‘ for cooling before molasses Water cooling of dispenser rods increases lifetime to 20‘‘ Enough for RF cooling Direct chip Z loading Trap lifetime around 1‘‘: Increase compression for faster thermalization => Necessity of constant currents (B fields)

RF : Fine tuning RF process was not reliable RF external antenna removed presence of resonances irregular input power/output power effect on the atoms not clear RF connected to Cu-Z clear dependence on frequency and power high reproducibility of results still room for improvement [  T/  n] TODO Tune frequency & power sweeps at the moment using linear sweeps, in 10‘‘ to 20‘‘ RF(end): 230kHz 190kHz 180kHz Step: 500kHz

Experiments in HD

The side guide Single wire guide Position of minimum: r 0  I wire / B bias Gradient  B 2 bias / I wire B bias // to chip rotational asymmetry Double wire B bias  to chip Rotational symmetry Possibility of bent wires Guide atoms around the chip Is B bias necessary? Similar result with extra wires Trap volume decreases Thywissen et al. EPJD 7,361 (1999)

The spiral Guiding 7 Li Two wire guide –counterpropagting currents –more than 2 complete turns –lifetime not influenced by guiding –Proves access to chip in 2D Luo et al. In preparation 5 mm duration of sequence: 220ms

Gimpel, Diploma Thesis, 2002 Cassettari et al PRL 85, 5483 (2000) Splitting atoms: y x t Based on magnetic fields Study of interactions with coherent atom waves Different structures to overcome barrier/reflection Wide wire influence? Adiabaticity? Wildermuth, Diploma thesis (2002); Folman et al, Adv.At.Mol.Opt.Phys 48, 263 (2002) BbBb IWIW IWIW 14 G44 G Bias field 7 Li

Another player: E U(r) = g F m F µ B B(r) – ½  E(r) 2 Electrostatic interaction does not depend on m F always attractive For realistic fields  constant Typical orders of magnitude ( 7 Li) U B [µK]  67 B [G] U E [µK]  98 E 2 [V/µm] 300V

B + E traps Static traps Atoms flow freely along a side guide Turning on electric field creates potential traps thermal 7 Li trapped with 300 V Magnetic guide: 1.6 A, 44 G Electrode voltage: 300 V Trap height: d chip = 70 µm 7 Li atoms

B + E splitters, motors Dynamics Electric fields can be varied; linearly summed atomic ‘motors’ time separation Lifetime not changed by extra fields (hot atoms) P.Krueger, X. Luo, K. Brugger, A. Haase, S. Wildermuth, S. Groth, I. Bar-Joseph, R. Folman, J. Schmiedmayer, quant-ph/ (2003) PRA in print

To do

Surface effects Tight trap Down to 62  m Is there fragmentation? Atoms too hot, need to be cooler Distinguish between fragmentation and imaging interferences Atoms can be loaded to chip potentials, for example to 10µm guides (500mA/7G) at T~3µK Trapped atoms d~100µm Released to guide Bondings T~5µK d~ 62µm Klein, Diploma Thesis Li Wide wire Further study required

Matrices Combine wires Use effectively multilayer technology up to 5 gold layer ‘topological freedom’ for wires connections Realize m x n traps dipole / quadrupole traps store and process independently cold atoms Next chip will be multilayer light several microns traps Use Stark shift detune individual traps for high resolution addressability

Optical fibers can be integrated Define grooves on chip for ease of fiber positioning Fibre cavities 90° fibres Nanooptics Peak corresponds to 150 atoms in the cavity finesse µm gap cavity formed by dielectric mirrors at the outer fibre ends piezo stretched to tune cavity length Fibre cavity of finesse 110 formed by dielectric mirrors at the outer fibre ends. A gap of 5 microns is included. The cavity length is scanned using a piezo stretcher Bruce Klappauf, ORC

The group Main Scientists Jörg Schmiedmayer Peter Schmelcher Post Docs Mauritz Anderson Louw Feenstra Thomas Fernholz PhD Students Karolina Brugger Leonardo Della Pietra Sönke Groth Albrecht Haase Alexander Kasper Peter Krüger Igor Lesanovsky Jürgen Rösch Stephan Schneider Christoph vom Hagen Stephan Wildermuth Marco Wilzbach Diploma Students Mihael Brajdic Elmar Haller Christian Hock Sebastian Hofferberth Anke Klafs Michael Schwarz Johanna Simon Thorsten Straßel Claudia Wagenknecht Ganjun Zhu Ex members Christiane Becker Bastian Engeser Ron Folman Hartmut Gimpel Sebastian Haupt Dennis Heine Ji-il Kim Matthias Klein Xueli Luo Hans Mathée Markus Rückel Thorsten Schumm Review: R.Folman, P.Krüger, J. Schmiedmayer, J.Denschlag, C.Henkel, Adv. At. Mol. Opt. Phys. 48, 263 (2002) FASTNet