Physica Scripta

You may also like

- How to control decoherence and
Quantum entanglement: a fundamental concept entanglement in quantum complex
systems?
finding its applications V M Akulin, G Kurizki and D A Lidar

- Effect of depolarizing noise on entangled

photons
To cite this article: Anton Zeilinger 1998 Phys. Scr. 1998 203 Ruchipas Bavontaweepanya

- Sensemaking of wave–particle duality in a

scientific outreach minicourse: a study
based in self-regulation
Matheus Henrique Thomas Becker,
View the article online for updates and enhancements. Leonardo Albuquerque Heidemann and
Nathan Willig Lima

This content was downloaded from IP address [Link] on 18/05/2025 at 13:49

Physica Scripta. Vol. T76, 203È209, 1998

Quantum Entanglement : A Fundamental Concept Finding its

Applications
Anton Zeilinger*
Institut fu
r Experimentalphysik, Universita
t Innsbruck, A-6020 Innsbruck, Austria

Received November 4, 1997 ; accepted December 17, 1997

PACS Ref : 03.67.[a, 42.50.[p

Abstract While the EPR-Bohr discussion for a long time was con-
Entanglement, according to the Austrian physicist Erwin Schro
dinger the sidered to be merely philosophical, David Bohm in 1952 [4]
Essence of Quantum Mechanics, has been known for a long time now to be introduced spin-entangled systems and John Bell in 1964
the source of a number of paradoxical and counterintuitive phenomena. Of showed that such entangled systems, when measuring corre-
those the most remarkable one is usually called non-locality and it is at the lated quantities, actually lead to di†erent predictions in the
heart of the Einstein-Podolsky-Rosen Paradox and of the fact that
quantum mechanical case than if one assumes that the
Quantum Mechanics violates BellÏs inequalities. Recent years saw an emer-
gence of novel ideas in entanglement of three or more particles. Most properties of the system measured are present prior to and
recently it turned out that entanglement is an important concept in the independent of observation. While by now a number of
development of quantum communication, quantum cryptography and experiments have conÐrmed the quantum predictions [5È7],
quantum computation. First explicit experimental realizations with two or from a strictly logical point of view the problem is not
more photons include quantum dense coding and quantum teleportation.
closed yet, because some loopholes in the existing experi-
1. Introduction ments still make it at least logically possible to uphold a
local realist world view [8].
Immediately after the discovery of modern quantum mecha- In recent years entanglement has become a new focus of
nics it was realized that it contains novel counterintuitive activity in quantum physics, because of (1) immense techno-
features, as witnessed most remarkably in the famous dia- logical and experimental progress (2) the discovery of novel
logue between Niels Bohr and Albert Einstein [1]. While non-classical features in multi-particle situations [9] and (3)
Einstein initially tried to argue that quantum mechanics is development of quantum information physics which heavily
inconsistent, he later turned his argument towards aiming at draws on multi-particle entangled systems.
demonstrating that quantum mechanics is incomplete. In his
seminal paper [2] where Einstein È together with Podolsky 2. Quantum coding
and Rosen È considers quantum systems consisting of two
particles such that, while neither position nor momentum of Let us consider a digital encoding scheme which is equiva-
either particle is well deÐned, the sum of their positions, that lent to encoding yes/no answers to individual questions or
is their center of mass, and the di†erence of their momenta, the truth value of elementary propositions. Usually these are
that is their individual momenta in the center of mass expressed by the bit value ““1ÏÏ for yes and ““0ÏÏ for no. A good
system, are both precisely deÐned. It then follows that mea- coding scheme should then be such that ““0ÏÏ or ““1ÏÏ can be
surement of either position or momentum performed on, distinguished clearly. Physical realizations of such coding
say, particle 1 immediately implies for particle 2 a precise schemes will therefore use systems which have two well-
position or momentum respectively without interacting with deÐned states. In classical coding these could be, for
that particle. Assuming that the two particles can be example, the two positions of a switch as in a simple relay,
separated by arbitrary distances, EPR suggest that measure- they could be two di†erent voltage levels in an electronic
ment on particle 1 cannot have any actual inÑuence on par- circuit, two di†erent colours of beads, a knot in my hand-
ticle 2 (locality condition) ; thus the property of particle 2 kerchief, etc.
must be independent of the measurement performed on par- Considering quantum coding schemes it is then natural to
ticle 1. To them, it then follows that both position and associate two di†erent orthogonal states to ““0ÏÏ and ““1ÏÏ.
momentum can simultaneously be well deÐned properties of Given therefore a two-state system we have to choose a
a quantum system. basis in which we wish to encode our information. Then any
In his famous reply [3] Niels Bohr argues that the two two orthogonal base states can be used to implement ““0ÏÏ
particles in the EPR case are always parts of one quantum and ““1ÏÏ. Let us therefore designate the chosen orthogonal
system and thus measurement on one particle changes the basis states as o 0T and o 1T respectively. Evidently a novel
possible predictions that can be made for the whole system feature of quantum information is that it can be in a super-
and therefore for the other particle. This is usually inter- position of these two states. A general superposition is
preted as implying a non-local feature of quantum mecha- called a qubit.
nics, though that is by no means the only possible Considering then the encoding of a large number of bits
interpretation of the situation. of information it is evident that this could be done by
simply assigning each bit its own two-state system. In that
ÈÈÈ case the information will be encoded into orthogonal
* electronic mail address : [Link]=[Link] product states, e.g. for three bits
Physica Scripta T 76
204 A. Zeilinger

o 0T o 0T o 0T
o 0T o 0T o 1T
o 0T o 1T o 1T
o 1T o 0T o 0T
o 1T o 1T o 0T
o 1T o 1T o 1T. (1)
The corresponding, in general 2n, states form a complete
orthonormal basis for the n-qubit space. Yet, alternatively, Fig. 1. In quantum dense coding Bob, having access to one of the two
in such a space we could also choose very di†erent bases entangled qubits only, can actually encode two bits of information, because
which even could be entangled. A maximally entangled basis he can switch around in the full Bell basis. Alice can identify the informa-
tion encoded by performing a complete Bell-state measurement on the
for two independent particles, two qubits, is entangled pair.
1
o W`T \ (o 0T o 1T ] o 1T o 0T),
J2
dense coding. This is the property that in order to switch
1 from any one of the four Bell states to all other four it is
o W~T \ (o 0T o 1T [ o 1T o 0T), sufficient to manipulate only one of the two qubits while in
J2
the classical case one has to manipulate both. Thus, the
1 sender Bob (Fig. 1) can actually encode two bits of informa-
o U`T \ (o 0T o 0T ] o 1T o 1T),
J2 tion into the whole entangled system by just manipulating
one of the two qubits. Let us, for example, assume that we
1 start from the state o W`T then we can obtain o W~T by just
o U~T \ (o 0T o 0T [ o 1T o 1T). (2)
J2 introducing a phase shift of n onto, say, the second qubit,
o U`T is obtained by Ñipping the second qubit and the last
This is the so-called Bell basis. It is important to notice
state o U~T is obtained by a combination of both.
that here we can still encode two bits of information, that is
In order to read out this information the receiver, Alice,
we have four di†erent possibilities, but now this encoding is
needs to be able to identify the four Bell states, that is she
done in such a way that none of the bits carries any well-
needs a Bell-state analyzer. It can be shown that, in order to
deÐned information on its own. All information is encoded
identify all four Bell states, one needs some non-linear inter-
into relational properties of the two qubits. It thus follows
action between the two qubits. In our experimental realiza-
immediately that in order to read out the information one
tion [11] of the dense coding scheme it was possible by just
has to have access to both qubits. The corresponding mea-
using linear elements like beam splitters and polarizers to
surement is called a Bell-state measurement. This is to be
identify three of the four possibilities and thus to encode
compared with the classical case where access to one qubit
and identify log 3 \ 1,58 bits of information per photon
is simply enough to determine the answer to one yes/no 2
manipulated.
question. In contrast, in the case of the maximally entangled
basis access to an individual qubit does not provide any
information. 4. Photon statistics at a beam splitter and Bell-state analysis
Formally speaking Bell-state analysis is not a problem. All
3. Quantum communication and dense coding
you have to do is project any incoming state onto the Bell
Whenever, say, two parties A (Alice) and B (Bob) wish to basis and you will Ðnd out by repeating the experiment
communicate with each other they have to agree Ðrst on a with which probability the original state can be found in
coding procedure, that is they have to agree which symbol one of the Bell states. The experimental problem is that thus
means what. In classical coding the situation is very simple. far no complete Bell-state analyser exists in the laboratory
Restricting ourselves to binary information, that is to bits, for any kind of quantum systems. This is due to the fact that
we need some information carrier which has two states. A complete Bell-state analysis requires non-linear interaction
famous historic example from the American revolution was between the two qubits [12] which has not been realized so
when Paul Revere informed the Revolutionaries about the far.
path taken by the Royals by displaying one or two lamps in Interestingly, partial Bell-state analysis is possible exploit-
the steeple of Old North Church in Boston. In quantum ing the statistics of two qubits at a beam splitter. The basic
physics again we can have information encoding in a novel principle of that Bell-state analyser rests on the observation
way using entangled states and thus encode information that of the states in equation 2 one state is antisymmetric.
into joint properties of elementary systems. Then the ele- This is the o W~T state which clearly changes sign upon
mentary systems themselves do not carry any information. exchange between the two particles. The other three states
A Ðrst elementary case where this is clearly demonstrated is are symmetric. We thus observe that the qubit obeys fer-
quantum dense coding. mionic symmetry in the case of o W~T and bosonic sym-
The maximally entangled Bell basis of eq. (2) has a very metry in case of the other three states. Thus far we have not
important and interesting property which was exploited by identiÐed whether we use bosons or fermions in our experi-
Bennett and Wiesner [10] in their proposal for quantum mental scheme. In fact, the four Bell states could very well
Physica Scripta T 76
 Quantum Entanglements 205

be either those of fermions or those of bosons. This is

because the states written in equation 2 are not the com-
plete states but can be amended by a spatial state which
also could be symmetric or antisymmetric. Then, in the case
of bosons the spatial part of the wave function has to be
antisymmetric also for the W~ state and symmetric for the
other three, while for fermions this has to be just reverse.
Let us Ðrst consider two photons, clearly bosons, where
we assume that the Bell states above describe the polariza-
tion of the photons, that is, an internal degree of freedom.
Then, clearly, the total state of the two photons has to be
symmetric. For the case of the two particles incident sym-
metrically onto a beam splitter (Fig. 2), one from each mode
o aT and o bT, the possible external states are
Fig. 3. Coincidence counts behind a beam splitter as a function of the dif-
1 ference of arrival time of two photons in an internally symmetric entangled
oW T\ (o aT o bT [ o bT o aT),
A J2 state. For large time di†erences the two photons are distinguishable and
therefore they behave independently and a certain coincidence rate equal to
1 the classical rate is observed. For zero time di†erence there are no coin-
oW T\ (o aT o bT ] o bT o aT), (3) cidences because both photons always leave the beam splitter together in
S J2 the same output port.

where o W T and o W T are anti-symmetric or symmetric

A S
respectively. Because of the requirement of symmetry the
total two-photon states are
detectors placed on each side after a beam splitter. How can
o W`T o W T, we then identify the other three states ? It turns out that
S
o W~T o W T, distinction between o W`T on the one hand and o U`T and
A o U~T on the other hand can be based on the fact that only
o U`T o W T, in o W`T the two photons have di†erent polarization while
S
o U~T o W T. (4) in the other two they have the same polarization. Thus per-
S forming polarization measurements and observing the
We note that only the state anti-symmetric in external photons on the same side of the beam splitter one can deter-
variables is also anti-symmetric in internal variables. It is mine whether they are in the state o W`T or in one of the
this state which also emerges from the beam splitter in an states o U`T and o U~T. It should be remarked that a gener-
external anti-symmetric state. This can easily be found by alization of this procedure implies that any three orthogonal
assuming Ðrstly that the beam splitter does not inÑuence the maximally entangled states can be distinguished from each
internal state and secondly that at the beam splitter we have other in the same way, because by local unitary transform-
a phase shift of n/2 upon reÑection. Actually it can easily be ations one can perform rotations in the two-dimensional
seen [13] that the spatially anti-symmetric state is an eigen- Hilbert-space.
state to any beam splitter operator. In contrast, in all three Consider now the same experiment with fermions where
cases of the symmetric external state o WsT the two photons again the Bell states describe the internal states, for example
emerge together in one of the two outputs of the beam split- if the two qubits are entangled in spin. We then Ðnd that
ter. This behaviour of symmetric states can readily be
demonstrated even using photons in an unentangled inter-
nal state [14]. Figure 3 and 4 show the consequent experi-
mental behaviour [15] for the entangled internal state o W`T
and that for the internally antisymmetric state o W~T.
It is therefore evident that the state o W~T can clearly be
discriminated from all the other states. It is the only one of
the four Bell states which leads to coincidences between

Fig. 2. A beam splitter with two inputs a and b and two outputs c and d.
We assume that two particles are incident onto the beam splitter one from Fig. 4. Coincidence rate for two photons in the internally antisymmetric
each side. The particles can then arise in two ways one in each of the two entangled state. Here again for large time di†erences we obtain the classical
output ports, they are either both reÑected or both transmitted. For behaviour. When the time di†erence approaches zero we get twice the coin-
bosonic states these two possibilities interfere destructively while for fer- cidence rate implying that the two photons never leave the beam splitter
mionic states they interfere constructively. together in the same output port.

Physica Scripta T 76
206 A. Zeilinger

the four possible states now are own ? In the case of entanglement swapping this interpretive
problem is even more striking.
o W`T o W T,
A Let us, for simplicity of discussion, just consider the case
o W~T o W T, where we have sources that produce our two qubits in the
S
anti-symmetric state o W~T. This state has the unique feature
o U`T o W T,
A that it is anti-symmetric in any basis. Thus, in terms of its
o U~T o W T, (5) information content, the statement is that we know the two
A
qubits are di†erent whatever basis we choose1. We thus
because of the antisymmetry requirement of the total state.
know simply by the choice of preparation that in each of the
For fermions therefore only one of the states is spatially
two entangled pairs to be used in entanglement swapping
symmetric, the other three are spatially anti-symmetric. It
the two qubits are completely di†erent.
will thus be in only one of the cases, namely for o W~T that
We now have to discuss the information content of Bell-
both fermions will emerge together from the beam splitter.
state analysis. At Ðrst we note that, whichever states we
In all other three cases they will emerge from di†erent sides
would produce at the sources, a fair Bell-state analyser will
[16]. Yet, remarkably, it is again this state which can imme-
return any of the four possible answers with equal probabil-
diately be distinguished from the other three because of its
ity of 25%. That is, the action of the Bell-state analyser is
distinct symmetry properties.
such that it projects the two photons onto an entangled
state and, since in our case the two qubits are themselves
5. Entanglement swapping and the nature of quantum members of maximally entangled pairs and therefore carry
information no information, this has to happen with equal probability
for all four Bell states measured. In fact, the Bell-state mea-
Entanglement used to be considered as a consequence of the
surement does not reveal any information about any of the
fact that the entangled particles interacted in their past or
qubits emitted by one of the two sources nor any joint infor-
that they came from a common source. That this is too
mation about each source. Yet, what we gain is joint (or
restricted a view is witnessed by the concept of entangle-
relative) information about the two sources.
ment swapping [17]. In the simplest case of entanglement
Suppose, speciÐcally, that in a certain experimental run
swapping we take two entangled pairs (Fig. 5) and subject
we obtain the result o W~T for the Bell-state measurement.
two particles, one from each source, to a Bell-state measure-
We then know that qubits 2 and 3 have been projected by
ment. Then the other two particles which have never inter-
the measurement onto a state which is characterized by the
acted in the past and also did not come from a common
fact that these two qubits are di†erent in whatever basis.
source are projected into an entangled state.
Interestingly, and again most remarkably, this statement is
This experiment, whose realization is presently proceeding
even true as none of the two qubits themselves are yet well-
in Innsbruck [31], has some deep implications for the
deÐned in any basis and have no properties by themselves.
meaning of information in quantum physics. As mentioned
Now we are in a position to complete our chain of
above, entanglement between two quantum systems implies
reasoning. By the properties of source 1 we know that qubit
that at least some information is stored only in joint or rela-
1 and qubit 2 are di†erent. By the result of the speciÐc Bell-
tional properties of the two systems. In the case of maximal
state measurement we know that qubit 2 and 3 are di†erent
entanglement there is actually no information stored in any
and, Ðnally, from the property of source 2 we know that
individual particle, all information is carried by all the par-
qubit 3 and 4 are di†erent. Therefore, since our qubits are
ticles jointly. This fact in itself poses already a challenge for
deÐned in a Hilbert-space of dimension 2 only, we conclude
a naive realist position. How can it be that particles have
that qubits 1 and 4 also have to be di†erent in any basis.
well-deÐned joint properties È for example when we know
Therefore they emerge in the anti-symmetric state o W~T.
that their polarizations in a certain basis are equal È
Analogous reasoning can be built up for the other three
without carrying any properties by themselves ? For a Bell
possible Bell-state measurement results.
state each of the systems is completely unpolarized on its
This analysis supports the interpretation that quantum
mechanics is just a formulation of what can be said about
quantum systems very much in the spirit of Bohr [18]. The
question as to what ““reallyÏÏ the properties of these quantum
systems are is empty and devoid of any meaning.
We could go even one step further realising that in
quantum physics information has a novel quality which
goes beyond its nature in classical physics. In classical
physics it is natural to assume that the objects we observe

ÈÈÈ
1 We remark that for the three symmetric states we can make analogous
statements but the situation is slightly more complicated. It turns out in
the end that, since the Hilbert-space of the four Bell-states is four-
dimensional, we can encode two independent bits of information into
Fig. 5. In entanglement swapping two Bell sources I and II emit one pair these four states. Therefore the four states can be characterised by yes/no
of entangled qubits each. One then performs a Bell-state measurement on a answers to two distinct questions. These two questions are questions
pair consisting of one qubit from each source. This projects the two outer- about the identity of the two qubits in two di†erent bases conjugate to
most qubits which have never interacted onto an entangled state. each other.

Physica Scripta T 76
 Quantum Entanglements 207

exist prior to and independent of any measurement and the the case of entanglement swapping, Alice will obtain each
information we gain by measurement is an entity of second- one of the four possible answers with equal probability, that
ary quality while the objects are primary. In contrast it is is her original qubit and her qubit from the ancillary pair
suggestive that in quantum physics the situation is that will be projected onto any one of the four Bell-states each
information can be a primary quality and the objects with with probability 25%. We note again that this measurement
their properties have a secondary quality, they can be con- does not reveal any information, neither about the proper-
structed in retrospect after the information has been gained ties of the original qubit nor about the properties of the
in a measurement. This is consistent with BohrÏs Caveat ancillary pair. So Alice obtains one of four possible results,
that we should be very cautious about making mental pic- ““W`ÏÏ, ““W~ÏÏ, ““U`ÏÏ, or ““U~ÏÏ. She then broadcasts this infor-
tures of quantum systems. Such pictures are only allowed in mation, that is two classical bits, such that Bob can receive
the context of a speciÐc experimental setup. In fact, most of them. By now Bob is in posssession of a speciÐc state as a
the misunderstandings and misinterpretations of quantum consequence of AliceÏs Bell-state measurement. Performing
mechanics result from assuming that quantum objects have one of four unitary transformations depending on AliceÏs
properties prior to and independent of the speciÐc appar- speciÐc result Bob can transform his particle into the orig-
atus chosen to observe them. inal qubit. We also note that the original qubit disappears
during the Bell-state measurement, it looses its identity, and
thus BobÏs qubit is not a copy but really a teleported
6. Quantum teleportation and the non-locality of information
reappearance of the original.
A most remarkable application of the concept of entangle- A remarkable property of the scheme is that the four
ment is quantum teleportation. Consider Ðrst the problem. unitary operations Bob has to perform are completely unre-
Suppose that Alice has an object which Bob, who could be lated to the state Alice wants to teleport. This is because
anywhere, might need at a certain time. In classical physics AliceÏs Bell-state measurement does not convey any infor-
what she can do is perform many precise measurements on mation about the original state.
the object and send the information to Bob who then can The experiment can most easily be understood again on
reconstitute the object. Within classical physics the accuracy the basis of the fact that the meaning of entanglement is that
of this is only limited by the precision with which Alice can it represents relational statements between quantum
measure and by the technical abilities of Bob. Yet, we know systems. So, in the case of teleportation we have to use
that in the end any measurement will run into the limi- entanglement twice, Ðrstly when the ancillary pair is produc-
tations imposed by quantum mechanics. It is evident that ed and secondly in the Bell-state measurement. It thus turns
no measurement whatsoever performed by Alice can reveal out that the entanglement of the ancillary pair implies a
the full quantum state of the object. We therefore ask which well-deÐned relation between these two qubits without
strategy can Alice pursue in order for Bob to obtain the deÐning their individual properties, and the Bell-state mea-
object in its full quantum state when he needs it. surement implies a well-deÐned relation between the qubit
A strategy proposed by Bennett et al. [19] uses exactly the to be teleported and one of the ancillaries. Therefore, the
information-theoretic features of entanglement mentioned second ancillary, by the same chain of reasoning as above in
above. According to that protocol, Alice and Bob have to our discussion of entanglement swapping, is in a well-
share in advance an entangled pair of quantum states (Fig. deÐned relationship to the original qubits to be teleported.
6). Let us consider for simplicity that the object is simply a It is again this remarkable feature of information in
two-state system, a qubit. Then Alice and Bob share from quantum mechanics that we have a closed chain of logical
the beginning an ancillary entangled pair which for conve- reasoning about relational properties without any statement
nience we again consider to be in the state o W~T. Subse- on the properties of the individual systems.
quently Alice performs a Bell-state measurement on her The major problem in the experimental veriÐcation of
qubit and one of the two ancillaries. As discussed above, for quantum teleportation [20] is the Bell-state measurement of
two independently created quantum systems. This means
that these two qubits have to be measured such that their
identity is lost, that it is not possible to infer which detection
event refers to which source. It turns out that in the experi-
ment it is rather tricky to achieve this situation. As can be
seen in [21], it involves an elaborate application of a
quantum erasure technique.

7. Multi-qubit entanglement in quantum computation

The most complex application of quantum entanglement
discussed presently is quantum computation [22]. Quantum
computation heavily uses superposition of multi-qubit
systems, i.e. entanglement. Therefore it is natural that
Fig. 6. In quantum teleportation Alice performs a joint Bell-state measure- among the most basic paradigms in the Ðeld are the pro-
ment on her initial particle and on one of an entangled ancillary pair from
duction and analysis both of Bell states and of so-called
the Bell source. This then projects the other particle of the ancillary into a
state which Bob can bring into exactly the original state of the initial parti- GHZ states [23], i.e. three-qubit entanglements.
cle using one of the four unitary transformations according to AliceÏs Bell- From the point of view of the novel possibilities intro-
state measurement result. duced by quantum computation it is important to realize
Physica Scripta T 76
208 A. Zeilinger

that superposition of multi-qubit states makes massive happens information technology will by necessity be
quantum parallelism possible [24]. This is that fact that a quantum. The radically new features of quantum entangle-
quantum computer can produce the result of various di†er- ment let us expect that this will not just be a limitation but
ent inputs at the same time if the input state is a super- that it will open up completely novel possibilities for com-
position of the individual informations. Consider, for munication and computation.
example, that using standard binary encoding the most The intricate interpretive questions raised by the multiple
simple GHZ state can be interpreted as a superposition of use of entanglement in the examples given here and as dis-
two numbers cussed above underline once more the deep conceptual
problems of quantum reality. It is hard to believe that there
1 1
(o 000T ] o 111T \ (o ““0ÏÏT ] o ““7ÏÏT) (6) is a possibility to escape Niels BohrÏs verdict ““There is no
J2 J2 quantum world, there is only an abstract quantum descrip-
that is, it represents a superposition of the numbers ““0ÏÏ and tionÏÏ.
““7ÏÏ. Taking this as the input to a linear quantum computer
it will produce in the output a superposition of the results
corresponding to the two inputs. Acknowledgements
It is well known that a most important practical problem
The author thanks the Nobel Committee for its hospitality. This work was
a future quantum computer faces is decoherence due to
supported by the Austrian Science Foundation (FWF), project S65/04 and
coupling to the environment. Somewhat analogous to clas- by the US National Science Foundation (NSF), grant number PHY97-
sical computers decoherence can be viewed as a quantum 22614. I also wish to acknowledge the collaboration with all members of
analogue of classical noise and it can be overcome using my group at the University of Innsbruck.
redundant information. Therefore, quantum error correction
codes [25] utilize the possibility of encoding n qubit into the
states of a Hilbert-space of higher dimension N. One then
References
exploits the additional degrees of freedom to eliminate the
quantum noise. This, naturally, implies entangled qubits in a 1. Bohr, N., ““Discussions with Einstein on Epistemological Problems in
Hilbert-space of high dimension. Atomic PhysicsÏÏ in Albert Einstein : Philosopher-Scientist““ (Edited by
P. A. Schilpp) (The Library of Living Philosophers, Evanston 1949), p.
While the Ðeld of quantum computation is signiÐed pre-
200.
sently by development of a vast number of di†erent codes 2. Einstein, A., Podolsky, B., and Rosen, N., Phys. Rev. 47, 777 (1935).
and theoretical approaches, the status of experiments is to 3. Bohr N., Phys. Rev. 48, 696 (1935).
verify the individual elementary steps including two-qubit 4. Bohm, D., ÏÏQuantum Theory““ (Prentice-Hall, Englewood Cli†s, NJ,
quantum gates. It is presently completely open whether 1951) p. 614.
5. Freedman, S. J. and Clauser, J. S., Phys. Rev. Lett. 28, 938-941 (1972).
future development will utilize multi-photon states as dis-
6. Aspect, A., Grangier, P. and Roger, G., Phys. Rev. Lett. 47, 460 (1981) ;
cussed in this paper, quantum encoding into individual 49, 91 (1982) ; Aspect, A., Dalibard, J. and Roger, G., ibid. 49, 1804
atoms [26] or individual ions [27] in trap conÐgurations or (1982).
bulk nuclear magnetic resonance techniques [28]. Yet, the 7. Kwiat, P. G., et al. Phys. Rev. Lett. 75, 4337 (1995).
fundamental individual steps in all these techniques are all 8. Pearle, P. M., Phys. Rev. D2, 1418 (1970) ; Clauser, J. F. and Shimony,
A., Rep. Prog. Phys. 41, 1881 (1978).
the same from a quantum information science point of view.
9. Greenberger, D., Horne, M. A. and Zeilinger, A., ÏÏBellÏs Theorem,
Quantum Theory, and Conceptions of the Universe““, (Edited by M.
Kafatos) (Kluwer, Dordrecht, 1989) p. 69 ; Greenberger, D. M., Horne,
8. Concluding comments M. A., Shimony, A. and Zeilinger, A., Amer. J. Phys. 58, 1131 (1990).
10. Bennett, C. H. and Wiesner, S. J., Phys. Rev. Lett. 69, 2881 (1992).
There are several further applications of entanglement in 11. Mattle, K., Weinfurter, H., Kwiat, P. G. and Zeilinger, A., Phys. Rev.
information science. One of them is the use of EPR pairs in Lett. 76, 4656 (1996).
12. Michael Reck, unpublished.
quantum cryptography as proposed by Ekert [29]. There
13. Zeilinger, A., Bernstein, H. J. and Horne, M. A., J. Modern Opt. 41,
one actually uses a test of BellÏs inequality to Ðnd out 2375 (1994).
whether or not an eavesdropper has gained any access to 14. Hong, C. K., Ou, Z. Y. and Mandel, L., Phys. Rev. Lett. 59, 2044
the qubits sent. (1987).
Another concept utilizing entanglement is quantum com- 15. Oberparleiter, M., Diploma-thesis, University of Innsbruck, 1997,
unpublished.
putation in distributed networks [30]. This concept present-
16. Loudon, R., ÏÏCoherence and Quantum Optics VI““ (Edited by J. H.
ly is in the stage of ideas and has not yet found experimental Eberly and L. Mandel) (New York : Plenum, 1990), p. 703 ; Loudon,
veriÐcation. It would imply extensive use of quantum tele- R., ÏÏDisorder and Condensed Matter Physics““ (Edited by J. A. Black-
portation, of Bell-state measurements, of entanglement man and J. Taguena) (Oxford : Clarendon, 1991), p. 441.
swapping, and of multi-particle entanglements. 17. Zukowski, M., Zeilinger, A., Horne, M. A. and Ekert, A. K., Phys.
Rev. Lett. 71, 4287 (1993).
Taken together, all the examples given above demonstrate
18. Nagel, B., in ÏÏPossible Worlds in Humanities, Arts and Sciences““.
already now that quantum entanglement is one of the most (Proceedings of Nobel Symposium 65) (Edited by Sture Allen)
important novel concepts in information science. It is fur- (Gruyter, Berlin 1989), p. 388.
thermore not unreasonable to even expect future applica- 19. Bennett, C. H. et al., Phys. Rev. Lett. 70, 1895 (1993).
tions of entanglement in information technology itself. This 20. Bouwmeester, D. et al., Experimental Quantum Teleportation, Nature
390, 575 (1997).
is supported by the observation that the present develop-
21. Zukowski, M., Zeilinger, A. and Weinfurter, H., ÏÏFundamental Prob-
ment of information technology towards further and lems in Quantum Theory““ (Edited by D. M. Greenberger, A.
further miniturization will reach the single quantum limit Zeilinger) (Annals of the New York Academy of Sciences, 1995) vol.
sometime in the Ðrst decades of the next century. Once that 755, p. 91.

Physica Scripta T 76
 Quantum Entanglements 209

22. C. H. Bennett, Phys. Today 48, 24 (1995). 27. Monroe, C. et al., Phys. Rev. Lett. 75, 4714, (1995) ; Wineland, D., this
23. Bruss, D., Ekert, A., Huelga, S. F., Pan, J.-W. and Zeilinger, A., Phil. volume.
Trans. R. Soc. Lond. A (1997) in press. 28. Gershenfeld, N. A. and Chuang, I. L., Science 275, 350 (1997).
24. Shor, P. W., in Proc. of the 35th Ann. Symp. on the Foundations of 29. Ekert, A. K., Phys. Rev. Lett. 67, 661 (1991).
Computer Science, Los Alamitos, CA, (IEEE Press, New York 1994) ; 30. Cirac, I., Zoller, P., Kimble, J. and Mabuchi, H., Phys. Rev. Lett. 78,
Ekert, A., Jozsa, R., Rev. Mod. Phys. 68, 733 (1996). 3221 (1997) ; Bose, S., Vedral, V. and Knight, P. L., Phys. Rev. A57,
25. Shor, P. W., Phys. Rev. A52, 2493 (1995) ; Steane, A. M., Proc. Roy. 822 (1998).
Soc. Lond. A452, 2551 (1996). 31. Pan, J.-W., Bouwmeester, D., Weinfurter, H. and Zeilinger, A., Phys.
26. Hagley, E. et al., Phys. Rev. 79, 1, (1997). Rev. Lett. 80, 3891 (1998).

Physica Scripta T 76