Bose 2.2 User Manual Page 107

  • Download
  • Add to my manuals
  • Print
  • Page
    / 155
  • Table of contents
  • BOOKMARKS
  • Rated. / 5. Based on customer reviews
Page view 106
In contrast, if the final lattice depth is chosen above the stability threshold, i.e. in the
cases
U
= 12
.
6
E
R
and
U
= 8
.
2
E
R
, there is no visible decay of the atom numbers. Here,
the TOF images basically show the usual interference pattern of a coherent array of
condensates (see section 4.4.2): we observe the central zero-momentum component and
the two side peaks corresponding to the lattice recoil momenta
±
2
~k
lat
(with
k
lat
=
π/d
lat
).
However, the central density peak shows a deviation with respect to the usual shape
obtained in non-dipolar BECs: it develops a
d
-wave shape that is similar to the one
observed in a single collapsed dBEC [36], shown in section 6.1. Due to our limited imaging
resolution, the characteristic cloverleaf shape of the central peak becomes more clear if
we expand the condensate for longer times
t
tof
, as shown in Fig. 6.4. For both the stable
in-trap configurations, considered here, we do not expect any in-trap evolution of the
system. Therefore, the observed
d
-wave collapse of the zero-momentum component must
be triggered by the time-of-flight (TOF-triggered collapse). Such collapse behaviour is
fundamentally different from the interaction-induced collapse shown in Ref. [36]. There,
the collapse happened partly during the TOF because the ramp in the scattering length
has not converged yet to its final value due to the eddy currents.
Finally, in the case
U
= 6
.
3
E
R
, which is just below the stability threshold, we observe a
slow decay of the atom number with increasing
t
hold
, but also a
d
-wave density pattern
of the central cloud. The observed dynamics of the system hence shows features of
both an in-trap collapse, as well as a TOF-triggered collapse. Using numerical real-
time simulations, we will investigate the complex dynamics of the system in details in
section 6.2.3. Before that, however, we study the time scale of the collapse dynamics
by performing a quantitative analysis of the absorption images taken in the collapse
measurements.
y
z
Fig. 6.4, d-wave collapse of the zero-momentum component:
The images show
the remaining coherent atoms of a BEC that is released from a lattice of
depth 8
E
R
, for the two different expansion times
t
tof
= 9
ms
and
t
tof
= 14
ms
.
Due to the limited resolution of our imaging system, the
d
-wave shape of the
central density peak is more visible at long expansion times (field of view:
190 × 497 m
2
).
By integrating the optical density in the images shown in Fig. 6.3, we obtain the number
of remaining coherent atoms for each lattice depth as a function of
t
hold
. For further
processing, we calculate the remnant fraction, defined as the number of coherent atoms
normalized by the total number of atoms, i.e. including the thermal cloud. Normalizing
the data in such way removes partly the shot-to-shot fluctuations that occur due to the
variation of the total atom number in the experiment. We then extract an atom loss rate
107
Page view 106
1 2 ... 102 103 104 105 106 107 108 109 110 111 112 ... 154 155

Comments to this Manuals

No comments