print this page


Atom-chip experiments

permanent staff :
Chris Westbrook, Isabelle Bouchoule, Alain Aspect

members :
Julien Armijo, Thibaut Jacqmin, Sébastien Gleyzes, Karim El-Amili

former members :
Christine Aussibal, Hai Nguyen,Jérôme Estève, Thorsten Schumm, Jean-Baptiste Trebbia, Ronald Cornelussen, Carlos Garrido-Alzar.


equipe puce
Atom chip team : Carlos Garrido-Alzar,
Julien Armijo and Isabelle Bouchoule
puce
optopuce
opto-atom chip team : Karim el-Amili and
Sébastien Gleyzes





On atom-chip experiments, atoms are confined in micro-magnetic traps realised by current carrying micro-wires deposited on the chip. Atom-chips present several advantages as compared to standard macroscopic cold atoms trapping schemes. First, using standard lithography technology, wires of width of the order of a micro-meter can be fabricated. Such small structures enable realisation of very high magnetic gradiant and thus strong confinements. They allow fast evaporative cooling of atomic cloud and thus simplify considerably the experiment since a very good vacuum is not required. Moreover, they permit the realisation of highly anistrop gases where one or two degrees of freedom are frozen so that the gas behaves as a 1D (resp. 2D) gases. Second, a large variety of trapping geometries can be designed. In particular, These advantages already enabled important experiments, that open the route to further studies~: atom-chip interferometers were realised, 1D gases were studied, Josephson physics was investigated ...

Our team contribute by several aspects on the atom chip and cold atoms studies. First, we performed several technical studies, important for the atom chip community. More precisely, we studied the potential roughness of an atom guide and we have shown that this roughness is due to geometric deformation of the wire edges. We have shown that it is possible to get rid of the roughness using modulated current. We also investigated the thermal performances of atom-chip. Second, we contribute to the study of 1D Boses gases. For this, we used the very elongated geometries accessible on an atom-chip to realise quasi-1D clouds. We then studied density fluctucations of such gases. 

We also begin an experiment where an atom-chip will contain a micro optical cavity for detecting atoms. We aim at reaching the strong coupling regime in order to realise non destructive measurement of single atoms.

return to top of page



The AlN-based chip is glued on a cupper block.

Current is brought using insulating HUV wires. A strip (1mmx8mm) made of berrilium-bronze is soldered at the extremity of each wire. Once folded, it ensures the elasticity requiered to permit a good electrical contact. It is maintained in grooves realised in a piece made out of PEEK. 18 electrical connexions are avalaible on the chip.

Picture of the experimental apparatus. The chip is mounted at 45 degrees inside a cubic stainless steel vacuum chamber. Two pumps ensure good pumping speed : a Titanium pump (above the chamber) and an ion pump. A vapor of rubidium is produced by rubidium dispensers placed on the sides of the chip mount.
Setup for absortion imaging The objective close to the vaccum chamber is a Melles Griot 3 lenses objective, diffraction limited with a numerical aperture 0.19 (corresponding to an Airy function whose first zero is at a radius of 2.5 microns.). The objective close to the camera is a simple doublet, working with numerical aperture smaller than 0.06. Two images are seen : the cloud and its reflection from the chip.




return to top of page

There are limits to the miniaturization of the structures and for several reasons, large currents are needed in atom-chip experiments. The maximal current that can be carried by each wire is thus a crucial parameter determining the possibilities of an atom chip. This maximal current is determined by the dissipation of the heat generated in the wires. Up to now, atom chips have mainly been realized on Silicon (Si) wafers, a good thermal conductor widely used in microfabrication. But, as Si is semiconductor, an electrically insulating layer, generally SiO2, needs to be placed between the wafer and the metallic wires. Unfortunately, SiO2 is also a thermal insulator, and this layer is the main limitation to the removal of heat in Si-based atom chips. On the other hand, AlN is a substrate material that has been especially selected for being simultaneously a good electrical insulator and a good thermal conductor and more and more groups working on atom chips are now moving to AlN substrates. Since no thermal contact resistance between the wire and the substrate is expected, much better heat dissipation is foreseen. We have fabricated such chips and measured their thermal behavior. We have shown that the heating in AlN-based atom chips is only governed by heat diffusion in the substrate, unlike the Si-based atom chips. We give a simplified model that accounts well for the observed heating. In particular we identify different heating regimes and give analytical approximate expressions. These results are published in arXiv:0906.2880

Heating of a 200 micronmeter width wire runing a current of 5A. 

The wire is either laid on paper (upper curve) or laid on cupper in air (middle curve) or glued to the cupper block (lower curve). Theoretical curves are also shown. See arXiv:0906.2880 for more detail.


 
 
Short time heating of a 7 micronmeter width wire running 1.7A.

Lines are predictions of the theoretcial model (solid) and a simplified $2D$ model (dashed). The crosses are experimental data. See arXiv:0906.2880 for more details.


return to top of page

An important limitation of microtraps is the potential roughness obsreved by all the groups working with atom-chips. Wang, Lukin and Delmer proposed that this roughness comes from geometrical deformation of the micro-wires. We have shown that this is indeed the case in our experimental situation. To this end, the potential roughness produced by a current carrying microwire has been deduced from the measurements of the density distribution of an atomic cloud at thermal equilibrium. We also measure the wire roughness using a scanning electron microscope and we compute the expected potential roughness it induces. The result is found to be in good agreement with the measurement of the potential using the atoms. Detailed informations about those results may be found in the articles [1] and [2].

Imperfections of the edges of the wire. 

The magnetic guide is realised using an electroplated wire of width 50 microns and height 6 microns. In the SEM image of the wire taken from the side (a), one can see that the edge deviation function  f is roughly independent of x. (c) Spectral density of f extracted from SEM images taken from the top as in Fig.(b).

 
 
Rough potentials normalized to the current in the Z-shaped wire for different heights from the wire.
Solid lines : potentials measured using cold atomic clouds. Dashed lines : potentials calculated from the measured geometric roughness of the edges of the wire. The different curves have been shifted by $6\,\mu$K/A from each other and heights above the wire are indicated on the right.
return to top of page

Since roughness is propotional to the wirecurrent, a modulation of the wire current produces an alternating roughness. If the modulation frequency is large compared to the frequencies of tha atomic motion, the atomic motion does not have time to follow the fast potential modulation and the atoms are sensitive to a residual potential without roughness. We apply this technique to a wire guide produced by five current carrying wires as shown in the picture below. For DC currents, the potential prensent some roughness that is responsible for a fragmentation of cold atomic cloud (see figure below). For AC current, no fragmentation of the cloud is seen and the roughness potential is measured to be reduced by at least a factor 5.

principle wires
Principle of the method : a change of sign of the current inverts the potential. The time averaged potential presents no roughness. Experimental implementation : the atomic guide is realised with 5 current carrying wires

roughness suppression
Roughness supression : modulating the wire current at 50 kHz supresses the potential roughness

return to top of page

Using measurement of noise in absorption images, we have studied density fluctuations in a quasi one dimensional Rubidium 87 Bose gas at thermal equilibrium. We analyse the datas in a local density approximation and we thus extract density fluctuations for a given linear atomic density n, at a given temperature T.

Different regimes have been observed depending on the ratio T/Tn where Tn= hbar n(gn/m) 1/2, g being the interaction coupling constant between the atoms. At large temperature, interatomic interactions are negligible. The observed fluctuations are in quantitative agreement with a calculation for an ideal Bose gaz. Fluctuations are in excess compared to the shot noise level expected for uncorrelated atoms because of the Bosonic bunching effect. At low temperature, the measured fluctuations are strongly reduced compared to the ideal gas case. This reduction is in good agreement with a calculation for an interacting Bose gas in the quasi-condensate regime.

Our atoms are confined in a very elongated trap produced by an H-shape wire. The transverse oscillation frequency is about 2.8 kHz where as the longitudinal one varies from 5 to 15 Hz.

These results are presented in cond-mat/0510397 .

We take about 300 absorption images taken in identical conditions. >From each image, we obtain the longitudinal profile N(zi), where zi is the ith pixel of the camera. We then look, for each pixel, at the statistical fluctuations of N. A large contribution to those fluctuations is due to the photon shot noise and we carrefully substract it. For very high temperature we obtain atom number fluctuations tah are linear in the mean value < N> as expected for uncorrelated atoms.

The slope is smaller than the expected value 1 because of the finite optical resolution.


return to top of page



At smaller temperature, but still high enough so that T>> Tn , we have observed density fluctuations in excess to the shot noise level due to bosonic bunching.

Measured density fluctuations for low density gazes. The solid lines are calculations for an ideal Bose gaz. The temperatures are deduced independently from the profile of the gaz.


return to top of page



In denser/colder clouds, we have observed density fluctuations much smaller than expected for an ideal gaz. This phenomena is due to repulsive interactions between atoms : the gaz enters the quasi-bec regime and density fluctuations are well described by a Bogoliubov calculation.


return to top of page



 

A single atom detection scheme integrated on the chip would be a major step in atom-chip technology. One way to realise this is to use a high finesse cavity whose resonnant frequency would be sensitive to the presence of a single atom. The presence of an atom would be then monitored as a change of light transmission for a light resonnant with a cavity mode, or by a phase shift of a transmitted beam which is far detuned from the cavity mode. Such a detection scheme could be used to monitor the output of an integrated atom-interferometer. On the other hand, the possibility to couple atoms with high finesse cavity opens the possibility to realise quantum information schemes on an atom-chip. For such applications, we need to access the strong coupling regim, for which g2/(Γs Γcav)>> 1, g being the atom-cavity coupling, 1/Γs the life time of the excited atom in vacuum and 1/Γcav the life time of a cavity photon. This regim should enable the realisation of non destructive detection of a single atom. Furthermore, if two atoms are coupled to a same cavity in the strong coupling regime many proposals show that it is possible to entangle them, for example via an exchange of a cavity photon. This opens the possibility to realise a quantum logic gate. The strong coupling regime has been achieved experimentally using macroscopic cavities of very high finesse in the optical regime\cite{Meshede,Kimble}. In such cavities, detection of a single atom has been obtained\cite{a voir}. Using a microwave transition between Rydberg states, an even more reach regime where g is both large compare to Γs and Γcav has been realised\cite{}. In this experiment, a logic gate between two atoms has been demonstrated. Our project consists in using a cavity based on integrated wave guides on the atom chip.

The idea is to realise a cavity using microbabricated optical wave guides. Mirors would be realised by Bragg grattings. A cavity with a small beam area could thus be produced, which is an important parameter to achieve the strong coupling regime. To enable a coupling of the cavity with an atom, the optical guide would have a hole. If this hole is small enough, it should not deteriorate too much the finesse of the cavity. The atom would then be brought in that hole to see an important electric field. If possible, the optical guide will be realised in GaAlAs, enabling a simple control of the optical index via an electrical current. This way, the cavity could be tuned to resonnance with the atoms. The atoms will be brought in the cavity by a magnetic guide perpendicular to the cavity as shown in the figure below.

Drawing of the micro-cavity and of the atom magnetic guide. The cavity, realised by an integrated optical wave guide, is closed by Bragg mirrors. A hole inside this cavity enable the insertion of atoms.
return to top of page



  1. I. Bouchoule, J.-B. Trebbia and C. L. Garrido Alzar~:''Limitations of the modulation method to smooth wire-guide roughness'', Physical Review A 77, 023624 (2008)
  2. J.-B. Trebbia, C. L. Garrido Alzar, R. Cornelussen, C. I. Westbrook, I. Bouchoule,  Phys. Rev. Lett. 98, 263201 (2007).~:''Roughness suppression in an AC atom chip''
  3. J.B. Trebbia, J. Estève, A. Aspect, C.W. Westbrook and I. Bouchoule,Phys. Rev. Lett. {\bf 97}, p.250403 (2006)~:''Experimental evidence for the failure of a mean field approach in an elongated Bose gas in the 1D weakly interacting limit''
  4. J. Estève, J-B. Trebbia, T. Schumm, A. Aspect, C. I. Westbrook and I. Bouchoule, Phys. Rev. Lett. 96, 130403 (2006)~:''Observations of Density Fluctuations in an Elongated Bose Gas: Ideal Gas andQuasicondensate Regimes'' 
  5.  Jerome Esteve, Thorsten Schumm, Jean-Baptiste Trebbia, Isabelle Bouchoule, Alain Aspect, Christopher Westbrook, physics/0503112, Eur. Phys. J D 35, 141 (2005)~: ``Realizing a stable magnetic double-well potential on an atom chip'' 
  6. Thorsten Schumm, Jerome Esteve, Cristina Figl, J.B. Trebbia, Christine Aussibal, Dominique Mailly, Isabelle Bouchoule, Christopher Westbrook and Alain Aspect, Eur. Phys. J. D 32, p.171-180 (2005)~:''Atom chips in the real world : the effetcs of wire corrugation''.
  7. Jerome Esteve, Christine Aussibal, Thorsten Schumm, Cristina Figl, Dominique Mailly, Isabelle Bouchoule, Christopher Westbrook and Alain Aspect,  Phys. Rev. A, 70,p. 043629 (2004)~:~``The role of wire imperfections in micro magnetic traps for atoms''

return to top of page


Offers of the "manip"

All offers of the Atom Optics group

return to top of page