A Quantum gas microscope for detecting individual atoms in a Bose-Hubbard optical lattice
New possibilities for quantum simulations of condensed matter models
Ultracold atoms give the unique opportunity to experimentally realize and study increasingly complex many-body quantum systems. One approach is to employ large samples of ultracold atoms to carry out such simulations [1]. The opposite approach is to assemble quantum information systems with full control over all degrees of freedom, atom by atom, ion by ion [2]. In this work we have created a quantum gas microscope that bridges between these two worlds. Thousands of individual atoms are detected with near-unity fidelity on individual sites of a Hubbard regime optical lattice. In addition, the single site addressability can be used for creating arbitrary potential landscapes and for local atom manipulation. This novel approach opens many new possibilities for quantum simulations and quantum information applications.
Quantum gases in optical lattices have been used to study fundamental models of condensed matter physics, including bosonic and fermionic Hubbard models [3]. In Bose-Hubbard experiments, a Bose-Einstein condensate is adiabatically loaded into the lowest band of an optical lattice. As the depth of the lattice is increased, a quantum phase transition has been observed from a conducting phase of the atoms to an insulating phase. In the conducting phase, the atoms are completely delocalized over the lattice and exhibit superfluid behavior. In the insulating phase, the strong interactions between the atoms localize them onto the lattice sites and suppress density fluctuations. The highly correlated state of the atoms in this regime is called a Mott insulator.
So far, experiments have studied the different phases of the Bose-Hubbard model through bulk measurements of the cloud, such as measurements of phase coherence or compressibility. Studying microscopic properties of the phases are difficult because a typical optical system for imaging quantum gases has a limited resolution of about 10 lattice sites. High resolution systems are further limited by averaging over the layers of a three dimensional lattice.
In our experiment, we can optically access a degenerate gas of rubidium atoms in an optical lattice with high resolution (~500nm), allowing us to resolve individual lattice sites[4]. We have achieved this resolution by combining large numerical aperture optics, a 2D quantum gas geometry and techniques from solid immersion microscopy. The high resolution optics have allowed us to image tens of thousands of individual atoms on the sites of an optical lattice with 640nm site spacing with high fidelity (98%). We can lower the temperature of the cloud below the condensation point, to obtain 2D Bose-Einstein condensates with temperatures of ~5nK as seen below.
The figure on the left is an in-situ image of a thermal cloud where the individual atoms/sites can be clearly resolved over a large field of view of over 100 microns. We can lower the temperature of the cloud below the condensation point, to obtain 2D Bose-Einstein condensates with temperatures of ~5nK as seen below. In this image, the edges of the Thomas-Fermi profile are clearly seen, and the condensate is surrounded by some thermal atoms.
These images are taken by collection of scattered photons from the atoms. Our objective (NA=0.8) collects 20% of the photons scattered by the atoms. During the imaging, the atoms heat up, and to prevent them from diffusing, we freeze the density distribution by non-adiabatically turning on a very deep lattice. The atoms are lost only due to background gas collisions over a timescale of 40s. During this time, we can collect 10,000 photons from each atom per second.
In this real time movie, the atoms are frozen in place but their population slowly decays. [.avi][.mpeg]
In this real time movie, the depth of the lattice in the vertical direction is lowered and thermal hopping is observed. [.avi][.mpeg]
High resolution microscopy of quantum gases opens up many new avenues in probing and manipulating the atoms in these systems [5]. Instead of using bulk measurements to study the different quantum phases of the atoms, we can investigate the microscopic properties of the system. For example it should be possible to study the change in the distribution of the atom number on lattice sites as the phase transition is crossed. In the superfluid state, the distribution of site occupations is Poissonian, while in the insulating state, a fixed onsite occupation is expected. This property of the Mott insulator has not yet been verified through direct measurement.
Beyond studying the Bose-Hubbard model, the quantum gas microscope enables more general quantum simulations. The high resolution of the system can be used to detect high frequency spatial correlations/fluctuations of quantities including spin, which would enable direct detection of new phases of matter such as antiferromagnetic ordering in a magnetic system. On the other hand, the microscope can also be used to project almost arbitrary light landscapes on the atoms, allowing simulation of novel systems. In fact, the lattices we are currently using in our experiments are projected through the microscope, allowing us to quickly change the lattice geometry by changing the mask pattern that is used in the projection. Finally, the ability to manipulate atoms on individual sites allows preparation of quantum states out of equilibrium and can also be useful in quantum information.
[1] Bloch, I., Dalibard, J. &, W. Zwerger, Rev. Mod. Phys. 80, 885–964 (2008), Greiner, M. & Foelling, S., Nature 453, 736-738 (2008).
[2] Schrader, D. et al., Phys. Rev. Lett. 93, 150501 (2004), Blatt, R. & Wineland, D. J., Nature 453, 1008–1014 (2008), Gaetan, A., Miroshnychenko, Y., Wilk, T., Chotia, A., Viteau, M., Comparat, D., Pillet, P., Browaeys, A., Grangier, P., Nature Physics 5, 115-118 (2009), Urban, E. et al., Nature Phys. 5, 110–114 (2009).
[3] Greiner, M., Mandel, O., Esslinger, T., Hänsch, T. & Bloch, I., Nature 415, 39–44 (2002), Jordens, R., Strohmaier, N., Gunter, K., Moritz, H. & Esslinger, T., Nature 455, 204–207 (2008), Schneider, U. et al., Science 322, 1520–1525 (2008).
[4] Nelson, K., Li, X. & Weiss, D., Nature Phys. 3, 556–560 (2007), Karski, M. et al., Phys. Rev. Lett. 102, 053001 (2009) , Gericke, T., Würtz, P., Reitz, D., Langen, T. & Ott, H., Nature Phys. 4, 949–953 (2008), W. S. Bakr, J.I. Gillen, A. Peng, S. Foelling, & M. Greiner, Nature 462 74-77 (2009)
[5] Gemelke, N., Zhang, X., Hung, C. & Chin, C., Nature 460, 995–998 (2009)
Publications:
Quantum Gas Microscope detecting single atoms in a Hubbard regime optical lattice
Nature 462 74-77 (2009).
Two dimensional quantum gas in a hybrid surface trap
Physical Review A 80 021602(R) 2009
The Quantum Gas Microscope
PhD. Thesis of Jonathon Gillen