Since its early development in the beginning of the 20th century, the quantum theory has reached tremendous success in describing the microscopic world. As a ground breaking example this theory perfectly describes the behavior of single to a few atoms coupled to an electromagnetic (EM) field (2012 Nobel Prize). Despite this success, the full understanding of larger systems known as N body correlated systems still remains to be unveiled. This is particularly the case in low temperature electronic solid state systems for which the temperature is far below the Fermi temperature (T/TF<0.005). Such regime is still mostly understood but we expect that the dynamics of the system will be driven by quantum correlations.
Two-dimensional electron gases (2DEG) that consist of electrons moving in a single atomic layer belong to this class of un-understood N-bodies correlated systems. For the quantum physics and condensed matter communities, they represent a great interest that could enlighten the conduction properties of graphene, the behavior of cuprates and other high-Tc superconductors or the properties of edge states in topological insulators. In this framework, the development of ultra-cold atom control offers toy-model systems to study the quantum properties of matter. In our project, we develop such a type of 2D ultra-cold atom quantum simulator using a new, original and challenging hybrid quantum system made of ultra-cold atoms and near field nano-structured EM potentials. Compared to usual far-field simulators, our apparatus allows us to raise all the energy scale at play in the underlying phenomena (nowadays a limiting factor) and to engineer novel types of long range interactions and band diagrams that should unveil these complex quantum phases.
AUFRONS - Ultra-cold atoms in a nano-structured optical lattice
The experimental sequence
It's the heart of the experiment where a pressure of 10e-10 mbar is reached. The following describes the experimental sequence to get cold atoms:
2DMOT and 3DMOT: Rubidium atoms are pushed from the 2DMOT chamber to the main 3DMOT chamber. Then, atoms are trapped using a Magneto-Optical Trap using a quadrupole field gradient and 5 collimators for the cooling at -2.7Γ and repumper beams. We load about 3.2e9 atoms in 12 seconds.
CMOT: Then, we realise a Compressed MOT by detuning the cooling frequency to -10Γ and increasing the magnetic gradient.
MOLASSES: After that, comes Molasses where the gradient is set to zero and the detuning is increased to -21Γ. At this stage, we know the residual magnetic offset and correct it with the compensation coils in the three directions. At the end, the final temperature is about 40 μK.
OPTICAL PUMPING: atoms are typically prepared in the state F=1, mF=-1 by using repumper and depumper beams σ- polarised before loading them in the magnetic trap.
MAGNETIC TRAP: usually 1.4e9 atoms are trapped by rising abruptly the quadrupole field to 60 G/cm and then increasing adiabatically to 160 G/cm. The atoms are heated up to 130 μK during this process but we can proceed with the evaporation towards Bose Einstein Condensates.
RF EVAPORATION: 2.3 seconds of radiofrequency evaporation enables us to get 9e7 atoms at 35/40 μK.
HYBRID CROSSED DIPOLE TRAP: 2e7 atoms are loaded by decompressing the magnetic trap until gravity compensation, and 1064nm dipole traps are turned on to finish the evaporation optically.
BECs of 1e6 atoms are obtained after 3s evaporation:
Thermal distribution: Thermal cloud, T>Tc
Double structure: BEC and thermal cloud, T=Tc
The optical bench
All laser beams at 780nm comes from a laser diode whose the feedback is achieved by saturated absorption on an atomic transition of Rb87 and two others slave laser diodes that are phase locked using beats with the Master.
Those three lasers are distributed on a separable optical table to get all the necessary beams on the experiment.
On the picture, the three laser beams are inserted from the top and manipulated with waveplates, mirrors, polarised beam splitters, shutters and acousto-optic modulators (to be fast). They are fibre coupled and sent to the experiment.
The imaging system
Once our lattice and BEC have been set up, we want to image the position of each site of the lattice. In other words a sub-wavelength resolution is needed to overcome the diffraction limit.
To this end, two energy states of Rb87 are modulated using the Stark effect such that only a few atoms can be excited on resonance for a given laser frequency. Scanning in frequency the imaging laser enables us to scan in position different atoms in the lattice and then reconstruct the site positions.
We are implementing a setup to have an optical lattice at 1529nm (5P-4D transition) to repump a tunable volume of atoms and get sub-wavelength details of the BEC.
lattice spacing is 7.7 um