Skip to main content

Full text of "Expansion of a quantum gas released from an optical lattice"

See other formats

Expansion of a quantum gas released from an optical lattice 

F. Gerbier^'B S. Trotzky, 2 S. Foiling, 3 U. Schnorrbcrgcr, 2 J. D. Thompson, 2 A. Widera, 4 I. Bloch, 2 

L. Pollet, 5 M. Troyer, 5 B. Capogrosso-Sansone, 6 N. V. Prokof'ev, 5,6 ' 7 and B. V. Svistunov 6,7 

1 Laboratoire Kastler Brossel, ENS, UPMC, CNRS ; 24 rue Lhomond, 75005 Pans, France 
2 Institut fur Physik, Johannes Gutenberg- Universitdt, 55099 Mainz, Germany. 
3 Department of Physics, Harvard University, Cambridge, MA 02138, USA. 
^Institut fur Angewandte Physik, 53115 Bonn, Germany 
5 Theoretische Physik, ETH Zurich, 8093 Zurich, Switzerland 
''Department of Physics, University of Massachusetts, Amherst, MA 01003, USA 
7 Russian Research Center "Kurchatov Institute", 123182 Moscow, Russia 
(Dated: August 15, 2008) 

We analyze the interference pattern produced by ultracold atoms released from an optical lattice. 
Such interference patterns are commonly interpreted as the momentum distributions of the trapped 
quantum gas. We show that for finite time-of-flights the resulting density distribution can, however, 
be significantly altered, similar to a near-field diffraction regime in optics. We illustrate our findings 
with a simple model and realistic quantum Monte Carlo simulations for bosonic atoms, and compare 
the latter to experiments. 

PACS numbers: 03.75.Lm,03.75.Hh,03.75.Gg 

Experiments with ultracold quantum gases in optical 
lattices rely heavily on time-of-flight (ToF) expansion to 
probe the spatial coherence properties of the trapped gas 
E B B 0, 0, S, 0, H ■ When the phase coherence length is 
large compared to the lattice spacing, the post-expansion 
density distribution shows a sharp interference pattern 
with the same symmetry as the reciprocal lattice. As 
the phase coherence length decreases, e.g., on approach- 
ing the Mott insulator (MI) transition, the visibility of 
this interference pattern decreases accordingly [lj]. To 
obtain a more precise understanding beyond this quali- 
tative description, it is usually assumed that the density 
distribution 71tof(f) of freely expanding clouds provides 
a faithful map of the initial momentum distribution. 

In this Letter, we point out that, in general, the ToF 
distribution differs from the momentum distribution for 
finite time-of-flight, the latter being recovered only in 
the "far-field" limit t — > oo. Practically, the ToF and 
momentum distributions become identical after a char- 
acteristic expansion time tpp = mR$l c /h, which depends 
on the particle mass m, the coherence length l c , and the 
cloud size Rq prior to expansion. This time scale can be 
understood in analogy with the diffraction of a coherent 
optical wave by a periodic grating. Then, the character- 
istic £ff in the expansion problem exactly corresponds 
to the Fresnel distance in the diffraction problem. The 
far-field regime is typically reached when the coherence 
length is short, for example for a cloud in the MI regime, 
or a thermal gas well above the critical temperature. We 
show that for phase-coherent samples where a sizeable 
fraction of the atoms are Bose condensed, the far-field 
condition is usually not met for typical expansion times 
used in current experiments [H,[lS,S,S, 0,0,1- Exper- 
imental measurements and quantum Monte-Carlo simu- 
lations are used to demonstrate that this results in sub- 

stantial changes in the ToF distribution. We also discuss 
implications for the interpretation of the ToF images. 

We consider an ultracold boson cloud released from 
a periodic trapping potential with cubic symmetry, lat- 
tice spacing d = Al/2, and lattice depth Vq given in 
units of the single-photon recoil energy £r = /i 2 /2mA 2 , 
where Al is the lattice laser wavelength. In addition to 
the lattice potential, an "external" harmonic potential 
is present, due to both the magnetic trap and the opti- 
cal confinement provided by the Gaussian-shaped lattice 
beams 0, H[ • This external potential is responsible for 
the appearance of a shell structure of alternating MI and 
superfluid regions in the strongly interacting regime. 

The density distribution after expansion for a time t is 
usually expressed as a product (see, e.g., [Toj] ) . 

(777 \ ^ n 777 V 

^) |7i (k)| 2 5(k), with k=— , (1) 

where an envelope function wq is the Fourier transform 
of the on-site Wannier function wq) and the interference 
term is 

5(k) = £ e 


Here the operator a J, creates an atom at site r M . To assess 
the validity of the far- field approximation used in Eq. |T]), 
we quickly outline its derivation. Neglecting interactions 
during expansion (see below), the atomic field operator 
can be expressed in Schrodinger's picture as ^(r,t) = 
Y,r v W v (r,t)a v where W„(r,t = 0) = w (r - r„). After 
the cloud is released, the wavefunction W v evolves in free 

m(r— r,A \ 4 m ( r 




LOht^> 1, with ujl the oscillation frequency at the bottom 
of a lattice well. In the limit t — > oo, the dependence 


15 20 25 

Expansion time (ms) 

FIG. 1: (a) Momentum distributions for a one-dimensional 
lattice with parabolic distribution of the occupation numbers 
calculated using Eq. ([5]) (solid line, expansion time t = 20 ms, 
dashed line: expansion time t — 100 ms, dot-dashed line: 
expansion time t — > oo). (b) Evolution of the peak amplitude 
A with expansion time t/tpF. The dashed line shows the 
expected near-field scaling in one dimension, A oc t/t-pF- The 
number of sites is 2JVtf+1 = 61 for (a) and (b). (c) Evolution 
of the width of the diffraction peaks with expansion time. 
The width has been normalized to the separation between 
two adjacent diffraction peaks for convenience. The circles 
show the experimental measurements and the solid line a fit 
by a hyperbola oc 1/t, as expected in the near-field. 

on the initial site position r„ vanishes, and one recovers 
Eq. |T|). For finite t, this dependence can be neglected in 
the envelope function ll|, but not in the phase factor. 
We thus obtain a generalized interference term 


Note that experimentally one observes a column distri- 
bution integrated along the probe direction, Sj_(k^) = 
/ dk z \w(k z )\ 2 St(k). This is included in latter compari- 
son with experiments, but in the following we base our 
discussion on Eq. ^ for simplicity. 

A fruitful analogy can be made with the theory of opti- 
cal diffraction. The formation of the interference pattern 
results from the interference of many spherical matter 
waves emitted from each lattice site, with phase relation- 
ships reflecting the initial quantum state of the boson 
gas. We can exploit this analogy further by defining the 
equivalent of a Fresnel distance usually introduced in the 
theory of optical diffraction to estimate the importance of 
the quadratic phase factor oc — r 2 . Because the corre- 
lation function (fl* a v ) suppresses contributions from sites 
distant by more than the characteristic coherence length 
l c , we can estimate the magnitude of the quadratic phase 

and ~ m fi/*° near the cloud edge. Here Rq the char- 
acteristic size of the cloud before expansion. The most 
restrictive condition to apply the far-field approximation 
thus reads t t FF , with 




As an example, for a 87 Rb condensate with l c m Rq ~ 
30 d and a lattice spacing d ~ 400 nm, one finds t FF ~ 
100 ms, much larger than typical expansion times t ~ 20 
ms in experiments 12j. In contrast, a gas with short co- 
herence length (e.g., in the MI regime), with l c > d, will 
enter the far-field regime after a few ms. We stress that 
the quadratic Fresnel term is intrinsically non-local, as 
the dephasing between two particular points r M and r„ 
depends not only on their relative separation but also 
on their absolute positions. Although this has little ef- 
fect deep in the superfluid or in the MI phase, this casts 
serious doubts on the validity of a local density approx- 
imation to compute quantitatively the ToF distribution 
in regimes where the coherence length is intermediate be- 
tween the cloud radius and the lattice spacing. 

To illustrate the influence of Fresnel terms on the in- 
terference pattern, we consider a ID lattice with uniform 
phase and parabolic distribution of the occupation num- 
bers, (fijjfij/) = c^Cy with c M = v/l — (/i/iVxp) 2 - The ToF 
distribution is given by 

S t (k) 

(2N TF + iy 

JV T f 

l = -N-T 


with k = kd, Ntf = Ro/d — 30 the Thomas- Fermi con- 
densate size in lattice units, and where (3 = y/md 2 /ht. 
The normalization factor (2Ntf + l) 2 would give the 
peak amplitude if the filling factor were uniform. We 
plot in Fig. the distributions corresponding to t = 
20 ms~ 0.2iFF which shows a significant broadening of 
the distribution for short time of flight when compared 
to the asymptotic result. For longer expansion times 
t ~ £ff ~ 100 ms, the far-field approximation is recov- 
ered to a good approximation. 

Qualitatively, we expect from dimensional arguments 
that the peak width scales as (f3N TF ) 2 = t FF /t in the 
near-field, while approaching a constant value in the far- 
field. The peak height thus increases as {t/t-p F ) T> in T> 
dimensions. This is confirmed by the one-dimensional 
calculation shown in Fig. [TJd. This dependence provides 
a mean to check the importance of near-field effects ex- 
perimentally. For the measurement, a sample of roughly 
10 5 87 Rb atoms has been been prepared in a three- 
dimensional optical lattice with a de pth V y — 6 Er, and 

inEq. © as ^(r^ - r ») 



near the cloud center. 

subsequently released for expansion 14| . After record- 
ing a series of absorption images for different expansion 
times, the width of the interference peaks was extracted 
using a Gaussian fit to the images. We plot the results 


in Fig. [TJ;, normalized to the separation between two 
diffraction peaks for convenience. The data confirms the 
t FF /t scaling, indicating that the far-field asymptote is 
not reached even after the longest expansion time avail- 
able in the experiment. 

We now discuss briefly the effect of interactions on the 
expansion, and show that this is negligible compared to 
the finite ToF effect. When the cloud has just been re- 
leased from the lattice potential, each on-site wavefunc- 
tion W/j, expands independently with a characteristic ex- 
pansion time lo^ 1 , until t f» t* = yjh/ (lu^Er) where 
the wavefunctions expanding from neighboring sites start 
to overlap. At this time, in the usual situation where 
ujjJj* ^> 1, the local density has dropped dramatically by 
a factor (wj,i) -3 <C 1. Hence, the interaction energy con- 
verts into kinetic energy on the time scale of a few oscilla- 
tion periods only, and expansion becomes rapidly ballis- 
tic. The parameter controlling the importance of interac- 

tions is given by V = fa « y/te eg* (H)^ with U 
being the on-site interaction energy. For typical parame- 
ters, T) is small (for instance rj ~ 0.05 for Vq = 10 En and 
the experimental parameters of Q). Hence, we expect 
only small corrections to the non-interacting picture of 
ballistic expansion. This has been confirmed using a vari- 
ational model of the expanding condensate wavefunction 
[l~5| . This model predicts that the "Wannier" envelope 
expands faster as compared to the non-interacting case, 
which does not affect the interference pattern, and picks 
up a site-dependent phase factor formally similar to the 
Fresnel term discussed previously, but with a very weak 
prefactor rj <C 1 which has negligible influence in prac- 
tice. We conclude that interactions essentially contribute 
to the expansion of the on-site wavefunctions, without 
significant dephasing of the interference pattern. 

The discussion so far focused on fully phase-coherent 
systems, which only applies to the weakly-interacting 
regime at low lattice depths. To investigate how the 
interference pattern is affected for strongly interacting 
systems (i.e., on approaching the Mott transition and be- 
yond), we have performed large-scale three-dimensional 
quantum Monte Carlo (QMC) simulations accounting 
for the external trapping potential using the worm al- 
gorithm 1(3, 12 1 m the implementation of Ref. fl8| . The 

calculations were performed for N = 8 x 10 4 atoms, us- 
ing exactly the same parameters and system sizes (up to 
~ 200 3 ) as in the experiments reported in The simu- 
lation was done at low constant temperature T = J/ks, 
where J is the hopping amplitude. Although simulations 
at constant entropy would be closer to the experimental 
situation, the temperature turns out to be approximately 
constant in this parameter regime [l9j |. 

The ToF distribution calculated for finite and infinite 
expansion times are shown in Fig. [2] The simulations 
confirm explicitly the analysis made above: the interfer- 
ence pattern is strongly affected in the superfluid phase, 




FIG. 2: Results from Quantum Monte Carlo simulations. On 
the left column, we show a horizontal cut through the ToF 
distributions for a finite expansion time t — 14 ms (solid line), 
compared to a cut through the profile calculated for t — > do 
(dashed line). Units for n± are arbitrary. The insets show 
directly the two-dimensional ToF distributions for t = 14 ms. 
On the right column, we show the in-trap density profiles for 
reference. The lattice depths are Vb = 12-Er (&)d), 15Er 
(b,e) and 17Er (c,f), respectively. 

and the effect becomes less and less pronounced as the 
lattice depth is increased and the Mott transition crossed. 
Note finally that the Fresnel phase suppresses the contri- 
bution from the edges of the cloud, thus favoring the con- 
tribution of the central region to the ToF pattern. This 
is especially important when superfluid rings surround a 
central MI region with lower coherence [201 ] . 

The interference p_attern is often characterized by its 
visibility 0,11 US 


"ToF(k max ) ~ ™ToF(k m i n ) 

«ToF(k n 

n To F(k 1J 


(2-7T, 0) and k m i n e? = V2(n, it) to 
cancel out the Wannier envelope in the division. We first 
evaluate the sensitivity of V to the Fresnel phase by plot- 
ting in Fig. ((3]) two theoretical "benchmark" curves as- 
suming perfect experimental resolution (dashed and dot- 
dashed lines for t = 14 ms and t — ► oo, respectively). We 
find little difference between the two curves when T j J is 
kept constant and small. Indeed, the Fresnel terms only 
matter for systems with large coherence length, where the 
visibility is by construction very close to unity. We con- 
clude that a detailed investigation of the superfluid side 
of the transition is better achieved by directly measuring 


Lattice depth V (f ft ) 

This could eventually provide a way to investigate the 
physics near the quantum-critical point without "par- 
asitic" contributions coming from coexisting superfluid 

Simulations were ran on the Brutus cluster at ETH 
Zurich. We acknowledge support from IFRAF, ANR 
(FG), DFG, EU, AFOSR (IB), the Swiss National 
Science Foundation (LP), NSF grant PHY-0653183 
(BCS,NP,BS) and DARPA (OLE project). 

FIG. 3: Visibility of the interference pattern as defined in 
Eq. (|6]). The dashed and dot-dashed lines show the Quan- 
tum Monte Carlo result for infinite and finite (t — 14 ms) 
expansion times, assuming perfect experimental resolution. 
The solid line is computed for t — 14 ms accounting for finite 
experimental resolution. Note that the comparison between 
experiment and simulation is only qualitative, since the sim- 
ulations were performed at constant temperature T = J/Ub 
while the experiment was not. 

the ToF distributions, whereas the visibility is well-suited 
for short coherence lengths. 

We also compare in Fig. [3] the experiments reported in 
Q to the predictions of the QMC simulations (solid line). 
Here, we emphasize that apart from the Fresnel terms, an 
accurate comparison requires to account for the experi- 
mental resolution, which is limited by two effects. First, 
the signal was obtained by integration over a square box 
centered around the maxima or minima, the integration 
area being ps (0.11 x 2ir/d) 2 in momentum units. This is 
comparable to a typical peak area, so that the visibility 
is calculated from the peak weight, rather than from its 
amplitude. Second, the finite resolution of the imaging 
system (about 6 /im) is not negligible for the sharpest 
peaks. Accounting for these two effects when evaluating 
the QMC data, we find good agreement with the exper- 
imental results. This entails that the experimental data 
arc compatible with the system remaining at low enough 
temperatures to cross a quantum-critical regime, in con- 
trast to the analysis made in Rcfs. 

21. 22] which included 

neither near-field expansion nor experimental resolution. 

In conclusion, we have analyzed the interference pat- 
tern observed in the expansion of a bosonic quantum gas 
released from an optical lattice. We showed that due to 
an additional Fresnel-like phase appearing for finite time 
of flight, the ToF distribution can be markedly different 
from the momentum distribution for clouds with large co- 
herence lengths. Conversely, the visibility as calculated 
from Eq. ^ is rather insensitive to this effect. 

The Fresnel phase acts as a magnifying lens for the 
central region undergoing a Mott insulator transition by 
suppressing the contribution of the outer regions of the 
cloud when the central density is close to integer filling. 

* Electronic address: 
[1] M. Greiner, O. Mandel, T. Esslinger, T. W. Hansen, and 

I. Bloch, Nature 415, 39 (2002). 
[2] T. Stoferle, H. Moritz, C. Schori, M. Kohl, and 

T. Esslinger, Phys. Rev. Lett. 92, 130403 (2004). 
[3] F. Gerbier, A. Widera, S. Foiling, O. Mandel, T. Gericke, 

and I. Bloch, Phys. Rev. Lett. 95, 050404 (2005). 
[4] F. Gerbier, A. Widera, S. Foiling, O. Mandel, T. Gericke, 

and I. Bloch, Phys. Rev. A 72, 053606 (2005). 
[5] S. Ospelkaus, C. Ospelkaus, O. Wille, M. Succo, P. Ernst, 

K. Sengstock, and K. Bongs, Phys. Rev. Lett. 96, 180403 


[6] K. Giinter, T. Stoferle, H. Moritz, M. Kohl, and 

T. Esslinger, Phys. Rev. Lett. 96, 180402 (2006). 
[7] I. B. Spielman, W. D. Phillips, and J. V. Porto, Phys. 

Rev. Lett. 98, 080404 (2007). 
[8] J. Catani, L. De Sarlo, G. Barontini, F. Minardi, and 

M. Inguscio, Phys. Rev. A 77, 011603(R) (2008). 
[9] M. Greiner, Phd thesis, Ludwig-Maximilian University 

Munich (2003). 

[10] P. Pedri, L. Pitaevskii, S. Stringari, C. Fort, S. Burger, 
F. S. Cataliotti, P. Maddaloni, F. Minardi, and M. In- 
guscio, Phys. Rev. Lett. 87, 220401 (2001). 

[11] Replacing wo ^fc = m ^ r "^ by wo (fc — ^j) requires 

t 3> -Ro/dojL, Ro/dy/uhE-Ei/h, or t > 1 ms for param- 
eters as assumed in the text. 
[12] This was independently established in [hi, although the 

impact on the ToF distribution was not analyzed. 
[13] E. Toth, A. M . Rey, and P. B. Blakie, e-print 

larXiv:0803.2922V 2 (2008). 
[14] The lattice beams used for this particular measurement 

have wavelengths X x — 765 nm for one axis x, and 
2 = 843 nm for the other two axis. The lattice depths 

are chosen such that the tunneling along the z-direction 

is equal to that along y, z. The experiment is otherwise 

identical to that described in 
[15] V. M. Perez-Garcia, H. Michinel, J. I. Cirac, M. Lewen- 

stein, and P. Zoller, Phys. Rev. Lett. 77, 5320 (1996). 
[16] N. V. Prokof'ev, B. V. Svistunov, and I. S. Tupitsyn, Zh. 

Eksp. Theor. Fiz 114, 570 (1998). 
[17] N. V. Prokof'ev, B. V. Svistunov, and I. S. Tupitsyn, 

Phys. Lett. A 238, 253 (1998). 
[18] L. Pollet, K. V. Houcke, and S. M. A. Rombouts, J. 

Comp. Phys. 225/2, 2249 (2007). 
[19] L. Pollet, C. Kollath, K. V. Houcke, and M. Troyer, New 

J. Phys. 10, 065001 (2008). 
[20] G. G. Batrouni, V. Rousseau, R. T. Scalettar, M. Rigol, 

A. Muramatsu, P. J. H. Denteneer, and M. Troyer, Phys. 


Rev. Lett. 89, 117203 (2002). [22] Y. Kato, Q. Zhou, N. Kawashima, and N. Trivedi, Nature 

[21] R. B. Diener, Q. Zhou, H. Zhai, and T.-L. Ho, Phys. Rev. Physics, doi:10.1038/nphys983 (2008). 

Lett. 98, 180404 (2007).