Optical Lattices, Ultracold Atoms and Quantum Information Processing
Optical Lattices, Ultracold Atoms and Quantum Information Processing
Information Processing
D. Jaksch
February 1, 2008
Abstract
We review novel methods to investigate, control and manipulate
neutral atoms in optical lattices. These setups allow unprecedented
quantum control over large numbers of atoms and thus are very promising for applications in quantum information processing. After introducing optical lattices we discuss the superfluid (SF) and Mott insulating (MI) states of neutral atoms trapped in such lattices and investigate the SF-MI transition as recently observed experimentally. In the
second part of the paper we give an overview of proposals for quantum information processing and show different ways to entangle the
trapped atoms, in particular the usage of cold collisions and Rydberg
atoms. Finally, we also briefly discuss the implementation of quantum
simulators, entanglement enhanced atom interferometers, and ideas for
robust quantum memory in optical lattices.
Introduction
The properties of a BEC described above make it an ideal tool for cold
atom physics and quantum state engineering by quantum optical methods.
One of the most promising experiments for future applications is the loading
of a BEC into an optical lattice [4, 5, 6]. Optical lattices are periodic conservative trapping potentials which are created by interference of travelling
laser beams yielding standing laser waves in each direction (see Fig. 1). The
laser light induces an AC-Stark shifts in atoms and thus acts as a conservative periodic potential. The usage of a BEC for loading has the advantage
that its large atom density allows a filling of few particles per site n & 1
while laser cooled atoms loaded into an optical lattice typically only allow a
filling factor smaller than one. Furthermore the atoms loaded from a BEC
are ultracold at temperatures very close to zero so that they practically behave as if their temperature was T = 0, in particular all of them occupy the
lowest Bloch band. The basic setup of optical lattices and the loading of a
BEC will be described in detail in Sec. 2.
An important novel feature which comes about because of the large filling
factor n & 1 is that interatomic interactions - due to swave collisions - of
two or more atoms occupying the same lattice site become important. The
interatomic interaction potential, which usually leads to incoherent collisions
in a thermal cloud of atoms and is responsible for the mean field in a BEC,
causes a coherent energy shift U of two particles occupying a single lattice
site. For the most frequently used alkali atoms 87 Rb and 23 Na the interaction
is repulsive U > 0 and thus the interaction energy competes with the kinetic
energy J gained by a particle when hopping from one lattice site to the next.
Both of these parameters depend on the laser intensity. When increasing the
laser intensity the barriers between neighbouring sites increases in height and
2
J goes down. At the same time two particles sitting in one site are pushed
together closer resulting in an increased interaction energy. Therefore the
competition between kinetic and interaction energy can be controlled by the
laser intensity which is an external easily changeable control parameter. In
Sec. 3 we will describe the physical consequences of different ratios J/U and
investigate the properties of the resulting ground states.
It turns out that the nature of the resulting quantum ground states
allows for a number of applications ranging from the controlled creation
of molecules [7], the study of strongly correlated quantum states [6], enhanced atom interferometry experiments and precision measurements [27],
controlled entanglement creation, to quantum computing [8, 9, 10, 12, 13].
In addition to having well defined initial motional states a number of different internal (hyperfine) atomic states can be trapped which results in additional well controlled degrees of freedom. Optical lattices are very versatile,
easily altered by changing the laser setup and the control over the quantum state of the atoms by quantum optical techniques is unprecedented.
Furthermore, decoherence times in optical lattices are large compared to
typical experimental times which allows the study of Hamiltonian dynamics of strongly correlated quantum states with time dependent parameters
and enables quantum information processing. All of these features make
optical lattices a unique system to study. We will discuss some of the possible applications in Sec. 4 focussing on applications in quantum information
processing. We will show how to implement quantum memory and quantum gates in optical lattices and review ideas for using optical lattices as a
quantum simulator for complex strongly correlated systems.
The high degree of precision that can be obtained in manipulating quantum optical systems and the large degree of isolation from the environment has led to a number of different quantum optical setups which might
be suitable for quantum computing. Several experiments in the groups of
D. Wineland [15] and R. Blatt [17] on ion traps have convincingly demonstrated two qubit gates. The next major step in these efforts is to scale them
to a larger number of ions which will probably be on the order of 30 [16].
Also, photons in an optical cavity might be used to implement a quantum
computer [14]. The biggest advantage of cavity systems is that they might
allow to couple resting qubits (atoms) to flying qubits (photons) and thus
enable the implementation of a quantum network.
There are also several non-quantum optical systems that have been identified as candidates for implementing a quantum computer. At the moment the most advanced system is NMR (Nuclear Magnetic Resonance)
[18]. However, although NMR systems can be used to perform demonstrations of quantum computing algorithms on very small systems it is as yet
not known how to scale an NMR quantum computer. Other promising candidates are superconducting Josephson junction arrays [19], quantum dots
[20] and somewhat more exotic systems like e.g. ammonia molecules con3
Figure 2: Interaction of a laser with a single atom. The red (blue) arrows
indicate red (blue) detuned lasers.
fined inside fullerenes [21]. Which of these setups will eventually lead to
a successful implementation remains to be seen. However, as we will show
for optical lattice setups in the remainder of this paper a lot of interesting
quantum physics is contained in the above systems. Therefore, even if some
of them will not succeed in performing scalable quantum computations, it
is certainly worth investigating them to learn more about their astonishing
quantum features.
2.1
Optical lattices
Let us for simplicity consider a two-level atom with an internal (hyperfine) ground state |gi and an excited state |ei in one spatial dimension
only coupled to a monochromatic classical laser field with a detuning as
schematically shown in Fig. 2. The laser is assumed not to cause significant
population in the excited state |ei and therefore we may neglect spontaneous emission from it. In a frame rotating with the laser frequency the
Hamiltonian (with ~ = 1) for the atom is given by
Hatom =
(
x)
p2
+ |ei he|
(|ei hg| + |gi he|) .
2m
2
(1)
Here x
and p are the coordinate and momentum operators respectively, and
the mass of the atom is m. The laser drives the transition between the
two atomic states with a Rabi frequency which is proportional to the
laser field and the dipole transition matrix element of the atom. We now
4
write
the atomic wave function as |(t)i = |e (t)i |ei + |g (t)i |gi where
e(g) (t) = he(g)| (t)i and derive the equations of motion for the wave
functions |e (t)i and |g (t)i. We find
d |g (t)i
dt
d |e (t)i
i
dt
p2
(
x)
|g (t)i
|e (t)i ,
2m
2
p2
(
x)
=
+ |e (t)i
|g (t)i .
2m
2
=
(2)
(3)
(x)2
.
4
(5)
We will always assume the atom to be interacting with a standing light wave
which arises from the interference of two counter propagating laser beams.
This means that the spatial dependence of the Rabi frequency is given by
(x) = 0 sin(kx),
(6)
(7)
T =
2m
5
2.2
Imperfections
b
eff
T ,
(8)
4
where is the spontaneous emission rate from the excited atomic level.
Therefore spontaneous emission will be negligible for large detunings .
Indeed, it has been shown in an experiment by Friebel et al. [22] that the
spontaneous emission rate in far detuned optical lattices can be of the order
of several minutes.
Since we are interested in having a few atoms per lattice we have to
consider the effects of light induced inelastic collisions between two atoms.
In Fig. 3 we have schematically shown the interaction potentials for the
different internal states (S + S) and (P + S) of two atoms approaching each
other at a distance r. As a consequence of this interaction the detuning
of the laser field changes as the particles come close to each other and
becomes = 0 at the so called Condon point RC . The probability for the
two particles to be excited to the states P + S increases around RC . The
atoms can then either form a quasistable molecule (for red detuning) which
decays by spontaneously emitting a photon or they can be promoted to an
unbound S + P state with a large kinetic energy equal to the detuning of
the laser at r = . Both of these effects should be kept as small as possible
to minimize loss from the optical lattice. There are several ways to do so:
Avoid a Condon point by detuning far to the red. However, this will
lead to a larger spontaneous emission rate compared to the case of a
blue lattice.
Avoid a red detuning that puts the Condon point close to a quasi
bound molecular state (shown in Fig. 3) as this enhances the probability of creating such molecules.
Choose the detuning such that the two potential curves have the
largest possible angle with each other at the Condon point. This makes
the probability of finding the two particles in the region where 0
small.
The above conditions for neglecting spontaneous emission and making laser
induced collisions unimportant can be fulfilled quite well experimentally and
6
then our above simple model for an optical lattice as a conservative potential
is valid.
2.3
Experimental setups
The situation where one pair of counter propagating laser beams interferes
and yields a trap potential in 1D while the atoms are only weakly trapped
by a magnetic field (or not trapped at all) in the other two dimensions is
usually called a 1D optical lattice. A BEC was first loaded into a 1D lattice
by B.P. Anderson and M.A. Kasevich [4]. In this experiment a magnetically
trapped BEC is loaded into a weak vertical optical lattice while at the same
time turning off the magnetic trapping almost completely. The experiment
is designed such that the interplay between the gravitational force and the
optical lattice leads to a peculiar interference pattern of atoms falling out
of the lattice which allows to study the coherence properties of the original
BEC.
7
The above setup with 3D lattice geometry was predicted to yield a qualitatively new behaviour of a dense cloud of ultracold atoms in 1998 [23]
and was experimentally realized in 2001 by M. Greiner et al. [6]. We can
gain some insight into the properties of the atoms trapped in the lattice by
looking at two limiting cases; the superfluid limit where the kinetic energy
dominates the repulsive interaction and the opposite limit which is called
the Mott insulating limit.
3.1
Superfluid limit
When the optical lattice is shallow the kinetic energy dominates over the
repulsion of two particles sitting in the same lattice site. In this situation
nearly all the atoms occupy the same single particle state. For very small
interaction energies this single particle state is very close to the Bloch wave
function with quasi momentum q = 0 of the lowest Bloch band, i.e. the single
particle ground state of the periodic potential. As this wave function is delocalized over the system there is long range off-diagonal coherence over the
whole lattice. States like this are very well described by the Gross-Pitaevskii
equation [2] which takes the interaction into account via a simple mean field
approximation. Neglecting interactions altogether we can understand the
form of the underlying Hamiltonian intuitively. We define a bosonic destruction operator al destroying an atom in the lattice site labelled l. Its
adjoint operator al correspondingly creates a particle in this lattice site and
nl = al al counts the number of particles in site l. By hopping (tunneling)
from one lattice site to an adjacent site an atom loses a kinetic energy J depending on the height of the barrier between the sites and the Hamiltonian
therefore reads
X
HSF = J
al am ,
(9)
hl,mi
where the sum runs over all pairs of nearest neighbours in the lattice. An
atom in the lattice will therefore minimize its energy when it is in a superposition state of many lattice sites.
A more detailed investigation shows that the state
!N
X
|SF i
al
|vaci
(10)
l
of N particles with |vaci the vacuum minimizes the energy. This state clearly
corresponds to each particle occupying the lowest Bloch state and spreads
the wave function of each particle over the whole lattice as indicated in
Fig. 5a. The state therefore shows large off diagonal long range coherence
as can be seen by correlating the destruction of an atom in lattice site l with
9
3.2
By increasing the lattice depth the kinetic energy goes down because the
barrier height increases. At the same time the size of each lattice site decreases which causes the atoms to repel each other more strongly. Again we
can understand this situation intuitively. We neglect any kinetic energy and
only consider the interaction energy of two particles in the same lattice site.
As usual the two particle interaction energy increases with the square of the
particle density which itself is proportional to the square of the number of
10
UX
nl (nl 1).
2
(11)
Here the parameter U is the interaction energy of two particles in the same
lattice site. Note the 1 term which is usually neglected in most mean
field theories. It ensures that a single particle does not gain any interaction
energy.
In this situation particles in the optical lattice minimize their energy by
staying away from each other instead of spreading out over the whole lattice.
Let us consider the case of commensurate filling M = N where the smallest
interaction energy is obtained by putting each particle into its individual
lattice site. A state like this is written as
Y
|MI i
al |vaci ,
(12)
l
is called a Mott insulator, and schematically shown in Fig. 5b. It does not
have any interaction energy hMI | Hint |MI i = 0 while the above SF state
would have an interaction energy of hSF | Hint |SF i = U N (N 1)/2M 2 .
Furthermore, the MI state has no long range coherence l,m = 0 for l 6= m
and also the particle number fluctuations are zero n2l = 0.
The MI state is much more stable than the SF state. The lowest lying
excitations are obtained by taking one particle out of its site and putting it
on top of an atom sitting in another lattice site. The energy required for
this is U independent of the size M , i.e. the system is characterized by a
finite gap U large compared to the small excitation energies in the SF limit.
Experimentally the situation of commensurate filling can be achieved
by suitably superimposing the optical lattice with a much weaker magnetic
trapping potential. The atoms fill those lattice sites which are in the center
of the wide trap and avoid the outer ones so that we obtain commensurate
filling in the center of the lattice. The magnetic trap acts as a soft boundary
which adjusts to the number of particles. Note, however, that the nature of
the MI will be altered close to this boundary.
3.3
Intermediate regime
Figure 7: Interference patterns for decreasing J/U (a-h). For U < Uc the
long range coherence leads to an interference pattern which clearly indicates
the band structure (a-d). The interference fringes vanish when the system
enters the MI regime U > Uc (e-h).
behaviour suddenly when the parameter g is changed over the critical point
gc . A quantum phase transition also occurs when the commutator of the
two parts of the Hamiltonian is not exactly zero but does not increase with
the size of the system and therefore can be neglected in the case of large
systems. This is exactly the situation we encounter in the so called BoseHubbard model with Hamiltonian HBHM = HSF + HMI we are considering
here for a changing ratio of J/U .
An optical lattice has the unique feature of permitting studies of the
dynamics of the quantum phase transition in HBHM , i.e., the parameters
U and J can be changed on a time scale much smaller than the coherence
time of the system but still sufficiently slow to avoid significant excitations.
The experiments carried out by M. Greiner et al. [6] to study this dynamics
started out from a Bose-condensate in a magnetic trap. Then the optical
lattice was (adiabatically) turned on. Its depth determines the ratio J/U and
thus the kind of ground state in the optical lattice. To measure the properties
of the atomic state the lattice potential is then turned off and the particles
are allowed to expand freely before their density distribution is measured.
In the case of a SF state which has long range off diagonal coherence the
particles interfere and the resulting interference pattern clearly indicates the
band structure of the lattice. When the system is in a MI state there is no
long range coherence and therefore the interference pattern disappears. This
behaviour is similar to interference in optics experiments where interference
patterns also vanish when there is lack of coherence. The corresponding
interference patterns as measured in [6] are shown in Fig. 7 for different
depths of the optical lattice.
Next we discuss some of the applications of MI states in optical lattices
focussing on applications in quantum computing. Afterwards we look at
13
The experiments on the SF to MI transition are a very important step towards realizing theoretical ideas for controlled entanglement creation in optical lattices by interactions between neutral atoms [8, 10, 9, 12]. Together
with single particle manipulations such entanglement operations constitute
the basic building blocks for a quantum computer [24] which fulfils all the
basic requirements as set out by D. DiVincenzo [25]. First important experimental steps towards realizing multi particle entangled states have recently
been achieved by Mandel et al. [26]. In this section we will first discuss
the requirements for implementing a quantum computer and then show how
they can be satisfied in an optical lattice.
4.1
1 X
1 X
|i |0i
|i |f ()i
L =1
L =1
(13)
in a single run. All the values of the function f are present in this superposition. However, we do not have access to all of this information since after a
measurement we will only obtain one result with a certain probability. The
property of using quantum superpositions to run the computer only once
was termed quantum parallelism by Feynman.
4.1.1
Requirements
1
X
j1 ,j2 ,...jN =0
1
X
j1 ,j2 ,...jN =0
(14)
These qubits can be in superposition and entangled states, which gives the
extraordinary power to the quantum computer; they also have to be well
isolated from the environment to prevent decoherence.
(ii) A universal set of quantum gates which allows the controlled manipulation of the qubits according to any unitary operation U so that
|i U |i. Fortunately, this task is enormously simplified given the fact
that any U can be decomposed as a product of gates belonging to a small
set, a so-called universal set of gates, i.e. if we are able to perform the gates
of this set we will be able to perform any unitary operation on the register
by applying a sequence of them. There are many sets of universal gates
and we will concentrate on the set containing one two qubit phase gate and
single qubit operations as this is the one most naturally arising in optical
lattice setups. Note that two-qubit gates require interactions between the
qubits, and therefore are the more difficult ones in practice.
(iii) Detection of the output state. One should be able to measure each
qubit in its computational basis. This process requires the interaction with
a measurement apparatus in an irreversible way and is thus not governed
by a unitary time evolution.
(iv) Initial state preparation. We must be able to erase quantum registers
to prepare the initial quantum state, for example the state |0, 0, . . . , 0i. This
is actually not an extra requirement, it is enough to detect the qubits and
to apply a single-qubit gate flipping the qubit |1i |0i and |0i |1i if
necessary.
(v) Scalability of the system. The difficulty of performing gates, measurements, etc., should not grow (exponentially) with the number of qubits.
Otherwise, the gain in the quantum algorithms would be lost.
(vi) In addition to the above we also require networking ability. The
static qubits used to perform quantum computations should be transformable into flying qubits which can easily be transmitted between different
locations.
15
4.2
In optical lattices two states of the ground state hyperfine manifold of the
trapped atoms |ai |0i and |bi |1i are ideal candidates to implement a
qubit. However, in order to be able to manipulate and control atomic qubits
it is necessary to know the position of each of the atoms precisely. While
random filling of optical lattices from laser cooled atoms and a superfluid
filling of a lattice (with large particle number fluctuations as discussed above)
do not provide sufficient knowledge on the position of the atoms a MI state is
ideally suited for this purpose. All the lattice sites are filled, i.e. each lattice
site contains a qubit and the fluctuations in the occupation numbers are
very small. For an appropriate choice of internal atomic states the optical
lattice allows to trap both of them and thus can hold a quantum register for
storing quantum information.
In principle it is straightforward to induce single qubit gates by using
Raman transitions between the two internal states |ai and |bi. Raman transitions with Rabi frequency R and detuning are described by the Hamiltonian
1
(15)
HR = (R |ai hb| + h.c.) + |bi hb| ,
2
which induce rotations of the qubit state on the Bloch sphere. The axis
and angle of this rotation depend on the choice of laser parameters and can
be chosen freely. The major problem in inducing single qubit operations
is the addressing of a single atom as it is difficult to focus a laser to spots
of order of an optical wave length which is the typical separation between
atoms in the lattice. Possible solutions to this difficulty are using schemes
for pattern loading [23, 30], where only specific lattice sites are filled with
atoms (cf. Fig. 8) or using additional marker atoms which specify the atom
the laser is supposed to interact with.
MI atoms in an optical lattice have already experimentally been used
as qubits and it has been shown that they support multi particle entangled
states [26]. The robustness of these qubits is, however, limited by stray
magnetic fields and spin echo techniques need to be used to perform the
experiments successfully. A different approach to obtaining robust quantum
memory uses a more sophisticated encoding of the qubits [27]. A 1D chain
16
with an even number of atoms encodes a single qubit in the states |0i =
|abab abi and |1i = |baba bai. These states both contain the same
number of atoms in internal states |ai and |bi and therefore interact with
magnetic fields in exactly the same way avoiding dephasing of the quantum
information stored in the chain. This and the fact that all the atoms have
to flip their internal state to get from one logical state to the other make
these qubits very robust against the most dominant experimental sources
of decoherence. However, at the same time this robustness makes it more
difficult to manipulate them. A laser pulse no longer corresponds to a simple
rotation on the Bloch sphere spanned by the logical states |0i and |1i which
makes the realization of a single qubit gate difficult. We will describe how
these problems can be circumvented in Sec. 4.5 and further details can be
found in [27].
17
4.3
Implementing a two qubit gate is more challenging than the single qubit
gates. The different schemes for two qubit gates can be classified in two
categories. The first version relies on the concept of a quantum data bus;
the qubits are coupled to a collective auxiliary quantum mode, like e.g. a
phonon mode in an ion trap, and entanglement is achieved by swapping the
qubits to excitations of the collective mode. The second concept which is the
basis for two qubit gates between atoms in optical lattices deploys controllable internal-state dependent two-body interactions. Examples for different
interactions are coherent cold collisions of atoms, optical dipole-dipole interactions [8, 9] and the fast two-qubit gate based on large permanent dipole
interactions between laser excited Rydberg atoms in static electric fields [10].
Besides these dynamical schemes for entanglement creation it is also possible
to generate entanglement by purely geometrical means [28]. We will now
discuss the different ways to achieve two qubit operations in optical lattices
4.3.1
The interaction terms introduced in Eq. (11) which describe swave collisions between ultracold atoms in one lattice site are analogous to Kerr
nonlinearities between photons in quantum optics. For atoms stored in optical lattices these nonlinear atom-atom interactions can be large [8], even for
interactions between individual pairs of atoms, thus providing the necessary
ingredients to implement two-qubit gates.
We consider a situation where two atoms in a superposition of internal
states |ai and |bi are trapped in the ground states of two optical lattice
sites (see Fig. 9a). Initially, at time t = these wells are centered at
positions sufficiently far apart so that the particles do not interact. The
optical lattice potential is then moved state selectively and for simplicity
we assume that only the potential for a particle in internal state |ai moves
to the right and drags along an atom in state |ai while a particle in state
|bi remains at rest. Thus the wave function of each atom splits up in space
according to the internal superposition of states |ai and |bi. When the wave
function of the left atom in state |ai reaches the second atom in state |bi
as shown in Fig. 9b they will interact with each other. However, any other
combination of internal states will not interact and therefore this collision
is conditional on the internal state. A specific laser configuration achieving
this state dependent atom transport has been analyzed in Ref. [8] for Alkali
atoms, based on tuning the laser between the fine structure excited p-states.
The trapping potentials can be moved by changing the laser parameters.
Such trapping potentials could also be realized with magnetic and electric
microtraps [29].
We therefore only need to consider the situation where atom 1 is in state
18
Figure 9: We collide one atom in internal state |ai (filled circle, potential indicated by solid curve) with a second atom in state |bi (open circle, potential
indicated by dashed curve)). In the collision the wave function accumulates
a phase according to Eq. (17). a) Configurations at times t = and b) at
time t.
|ai and particle 2 is in state |bi to analyze the interactions between the two
atoms. The positions of the potentials are moved along trajectories x
a (t)
b
and x
(t) = const so that the wave packets of the atoms overlap for a certain
time, until they are finally restored to the initial position at the final time
t = . This situation is described by the Hamiltonian
X (
p )2
x
x
(t) + uab (
+V
xa x
b ).
(16)
H=
2m
=a,b
Here,
x
a,b
and
pa,b are
position
and
momena,b
a,b
tum operators, V
x
x
a,b (t) describe the displaced trap potentials
and uab is the atomatom interaction term (which lead to the interaction
term in Eq. (11)). Ideally, we want to implement the transformation from
before to after the collision,
0a (xa x
a ( ))0b (xb x
b ( )) ei 0a (xa x
a ( ))0b (xb x
b ( )), (17)
where each atom remains in the ground state 0a,b of its trapping potential
and preserves its internal state. The phase = a + b + ab will contain a
contribution ab from the interaction (collision) and (trivial) single particle
kinematic phases a and b . The transformation Eq. (17) can be realized
in the adiabatic limit, whereby we move the potentials slowly on the scale
given by the trap frequency, so that the atoms remain in their motional
19
|0, 1i |0, 1i
|1, 0i |1, 0i
|1, 1i |1, 1i
(19)
(20)
with j (t, xj ) Rabi frequencies, and j (t) detunings of the exciting lasers.
Here, u is the dipole-dipole interaction energy between the two particles. We
have neglected any loss from the excited states |rij . We discuss two possible
realizations of two qubit gates with this dynamics. The most straightforward
way to implement a two-qubit gate is to just switch on the dipole-dipole
interaction by exciting each qubit to the auxiliary state |ri, conditioned on
the initial logical state. This can be obtained by two resonant (1 = 2 = 0)
laser fields of the same intensity, corresponding to a Rabi frequency 1 =
20
sgn() || 2 + 2
(21)
( ) =
dt sgn()
2
t0
with = 2 /(4 + 2u) the detuning including a Stark shift. For a specific choice of pulse duration and shape (t) and (t) we achieve ( ) = .
To satisfy the adiabatic condition, the gate operation time has to be
approximately one order of magnitude longer than in the other scheme discussed above. In the ideal limit j u, the dipole-dipole interaction energy
shifts the doubly excited state |rri away from resonance. In such a dipoleblockade regime, this state is therefore never populated during gate operation. Hence, the mechanical effects due to atom-atom interaction are greatly
suppressed. Furthermore, this version of the gate is only weakly sensitive to
the exact distance between the atoms, since the distance-dependent part of
the entanglement phase . For the same reason, possible excitations
in the particles motion do not alter significantly the gate phase, leading to
a very weak temperature dependence of the fidelity.
In principle the controlled implementation of the above single and two21
4.4
Figure 10: The Hofstaedter butterfly. Each dot in this fractal indicates an
eigenenergy of the atoms moving in a magnetic field with parameter as
given in the text. The energy is normalized to 0 which is given by the
modulus of the hopping energy between two sites.
A more detailed investigation shows that interesting spin-chain Hamiltonians like the (anti)ferromagnetic Heisenberg Hamiltonian can be simulated
without the need for local addressing of single atoms and could thus be
realized with standard optical lattice setups. The ability to perform independent operations on each of the qubits would allow all possible bipartite
Hamiltonians to be simulated and in particular the study of quantum phase
transitions in spin chains. We will next briefly study an optical lattice setup
with Rydberg interactions between adjacent atoms which shows a quantum
phase transition and could be viewed as a quantum simulator [27]. We will
also show how it can be used to implement an atom interferometer whose
sensitivity is enhanced by the created entanglement. Furthermore, the ideas
presented in the next section are the basis for implementing single qubit
gates on robust qubits as introduced in Sec. 4.2.
4.5
(j)
with x the Pauli x-matrix for the two states (viewed as a spin) of the
j-th atom and B > 0 the energy associated with this process. The ground
state of this Hamiltonian is obtained by putting each atom into an equal
(24)
accounting for the energy difference of having two adjacent particles in the
same vs. different internal states. In this case the ground state depends on
24
(26)
(27)
Both of these states are maximally entangled multi particle states, i.e. extensions of the well known GHZ states for three particles.
In the total Hamiltonian H = Hh + Hi the interaction energy and the
hopping energy compete with each other resulting in a quantum phase transition. When the interaction energy is kept constant and the hopping term
is switched off adiabatically, as e.g. shown in Fig. 11 the state of the system
will dynamically change from |h i which is a product state to one of the
two states |ir i or |ia i depending on the sign of the interaction. The exact
values of the parameters and depend on the details of the dynamics
and are discussed in [27]. These maximally entangled states can serve different purposes depending on the sign of W . Let us discuss the possible
applications of these two kinds of maximally entangled states
4.6
Repulsive interactions
In this case both parts of the superposition state have the same number
of particles in each of the two internal states and thus external stray fields
act identically on both, therefore not affecting the parameters and .
Because of this stability one can use them for storing quantum information
in a robust way as described in Sec. 4.2. Single qubit gates can be performed
by dynamically going back and forth through the quantum phase transition
in the whole chain changing the parameters and in a controlled way.
25
4.7
Attractive interactions
For attractive interactions the terms in the superposition of |ia i will respond to external fields very differently. Therefore the relative phase between the parameters and will be very susceptible to these fields, in fact
for N particles in the chain this phase will be N times larger than if there
was just a single particle. The two parts of |ia i can thus be used as two
arms of an entanglement enhanced atomic interferometer.
Summary
This article reviewed recent advances in controlling and manipulating neutral atoms by quantum optical methods concentrating on the use of optical
lattices. We investigated novel physical features arising from interactions
in ultracold dense clouds of atoms as provided by BEC and showed how
they enable experimentalists to gain control over the spatial degrees of freedom of individual atoms. We also studied possible new applications - in
particular in quantum information processing - which make use of this unprecedented quantum control. We showed that universal sets of quantum
gates can be implemented and how they might lead to a universal scalable
quantum computer in the future. Furthermore we also investigated possibilities for shorter term applications of optical lattices as quantum simulators
for strongly correlated systems.
References
[1] M.H. Anderson, J.R. Ensher, M.R. Matthews, C.E. Wieman, and E.A.
Cornell, Science, 269, 198 (1995); K.B. Davies, M.-O. Mewes, M.R.
Andrews, N.J. van Druten, D.S. Durfee, D.M. Kurn, and W. Ketterle,
Phys. Rev. Lett. 75, 3969 (1995); C.C. Bradley, C.A. Sackett, J.J.
Tollett, and R.G. Hulet, Phys. Rev. Lett. 75, 1687 (1995).
[2] For a review see, e.g.: J.R. Anglin and W. Ketterle, Nautre 416,
211 (2002); F. Dalfovo S. Giorgini, L.P. Pitaevskii, and S. Stringari,
Rev. Mod. Phys. 71, 463 (1999); L.P. Pitaevskii and S. Stringari, BoseEinstein condensation, Oxford University Press (2003).
[3] For further information see: http://www.nobel.se/physics/laureates/2001/.
[4] B.P. Anderson, and M.A. Kasevich, Science 282, 1686 (1998); C. Orzel,
A.K. Tuchman, M.L. Fenselau, M. Yasuda, and M.A. Kasevich, Science,
291, 2386 (2001).
26
J.I.
Cirac,
and
P.
Zoller,
[15] Q.A. Turchette, C.S. Wood, B.E. King, C.J. Myatt, D. Leibfried,
W.M. Itano, C. Monroe, and D.J. Wineland, Phys. Rev. Lett. 81, 3631
(1998); D. Leibfried, B. DeMarco, V. Meyer, D. Lucas, M. Barrett,
J. Britton, W.M. Itano, B. Jelenkovi, C. Langer, T. Rosenband, and
D.J. Wineland, Nature 422, 412 (2003).
[16] D. Kielpinski, C. Monroe, and D.J. Wineland, Nature 417, 709 (2002).
[17] S. Gulde, M. Riebe, G.P.T. Lancaster, C. Becher, J. Eschner,
H. Haffner, F. Schmidt-Kaler, I.L. Chuang, and R. Blatt, Nature
421, 48 (2003); F. Schmidt-Kaler, H. Haffner, M. Riebe, S. Gulde,
G.P.T. Lancaster, T. Deuschle, C. Becher, C.F. Roos, J. Eschner, and
R. Blatt, Nature 422, 408 (2003).
[18] L.M.K. Vandersypen, M. Steffen, G. Breyta, C.S. Yannoni, M.H. Sherwood, and I.L. Chuang, Nature 414, 883 (2001); H.K. Cummins,
27
C. Jones, A. Furze, N.F. Soffe, M. Mosca, J.M. Peach, and J.A. Jones,
Phys. Rev. Lett. 88, 187901 (2002).
[19] Y. Yu, S. Han, X. Chu, S.-I. Chu, and Z. Wang, Science 296, 889 (2002);
J. Martinis, S. Nam, J. Aumentado, and C. Urbina, Phys. Rev. Lett.
89, 117901 (2002); Yu.A. Pashkin, T. Yamamoto, O. Astafiev, Y. Nakamura, D.V. Averin, and J.S. Tsai, Nature 421, 823 (2003); T. Yamamoto, Yu.A. Pashkin, O. Astafiev, Y. Nakamura, and J.S. Tsai,
Nature 425, 941 (2003).
[20] X. Li, Y. Wu, D. Steel, D. Gammon, T.H. Stievater, D.S. Katzer,
D. Park, C. Piermarocchi, and L.J. Sham, Science 301, 809 (2003);
B.E. Kane, Nature 393, 133 (1998).
[21] A.J. Ferguson, P.A. Cain, D.A. Williams, and G.A.D. Briggs,
Phys. Rev. A 65, 034303 (2002).
[22] S. Friebel, C. DAndrea, J. Walz, M. Weitz, and T.W. Hansch,
Phys. Rev. A 57, R20 (1998).
[23] D. Jaksch, C. Bruder, J.I. Cirac, C.W. Gardiner, P. Zoller,
Phys. Rev. Lett. 81, 3108 (1998).
[24] D. Deutsch, Proc. Roy. Soc. Lond. A, 425, 73 (1989); D.P. DiVincenzo, Phys. Rev. A 50, 1015 (1995); H.F. Chau and F. Wilczek,
Phys. Rev. Lett. 75, 748 (1995); A. Barenco, Proc. Roy. Soc. Lond. A
449, 679 (1995); D. Deutsch, A. Barenco and A. Ekert,
Proc. Roy. Soc. Lond. A 449, 669 (1995).
[25] D.P. DiVincenzo, Fortschr. Phys. 48, 771 (2000).
[26] O. Mandel et al., Nature 425, 937 (2003).
[27] U. Dorner, P. Fedichev, D. Jaksch, M. Lewenstein, and P. Zoller,
Phys. Rev. Lett. 91, 073601 (2003).
[28] L.-M. Duan, J.I. Cirac, and P. Zoller, Science 292, 1695 (2001).
[29] T. Calarco, H.-J. Briegel, D. Jaksch, J.I. Cirac and P. Zoller,
J. Mod. Opt. 47, 2137 (2000).
[30] S. Peil, J.V. Porto, B. Laburthe Tolra, J.M. Obrecht, B.E. King,
M. Subbotin, S.L. Rolston, and W.D. Phillips, Phys. Rev. A 67,
051603(R) (2003).
[31] P. Shor, Proc. 35th An. Symp. On Foundations of Comp. Sci., IEEE
Press, Los Alamitos, CA, 1994.
[32] A. Srensen and Klaus Mlmer, Phys. Rev. Lett. 83, 2274 (1999).
28
29