LETTER doi:10.

1038/nature11472

Quantum teleportation over 143 kilometres using

active feed-forward
Xiao-Song Ma1,2{, Thomas Herbst1,2, Thomas Scheidl1, Daqing Wang1, Sebastian Kropatschek1, William Naylor1,
Bernhard Wittmann1,2, Alexandra Mech1,2, Johannes Kofler1,3, Elena Anisimova4, Vadim Makarov4, Thomas Jennewein1,4,
Rupert Ursin1 & Anton Zeilinger1,2

The quantum internet1 is predicted to be the next-generation stability is necessary. Moreover, the complexity and environmental
information processing platform, promising secure communica- requirements of a quantum teleportation set-up are increased signifi-
tion2,3 and an exponential speed-up in distributed computation2,4. cantly compared to previous two-photon experiments, which provides
The distribution of single qubits over large distances via quantum significant experimental and technological challenges. The work pre-
teleportation5 is a key ingredient for realizing such a global plat- sented in ref. 18 was a significant achievement in long-distance
form. By using quantum teleportation, unknown quantum states quantum communication, harnessing the entanglement between dif-
can be transferred over arbitrary distances to a party whose loca- ferent degrees of freedom of a single photon. However, because the
tion is unknown. Since the first experimental demonstrations of teleported qubit was not provided independently from the outside, the
quantum teleportation of independent external qubits6, an applicability of that scheme is limited. Most earlier experiments on
internal qubit7 and squeezed states8, researchers have progressively teleportation of qubits and squeezed states6–8 were in-laboratory
extended the communication distance. Usually this occurs without demonstrations, and hence the communication distances were rather
active feed-forward of the classical Bell-state measurement result, short. Although fibre-based teleportation has been demonstrated
which is an essential ingredient in future applications such as com- experimentally20,21, the maximum transmission distance is limited by
munication between quantum computers. The benchmark for a the intrinsic photon loss in optical fibre, unless quantum repeaters are
global quantum internet is quantum teleportation of independent involved22. In comparison to these previous studies, the experiment
qubits over a free-space link whose attenuation corresponds to the presented in this Letter achieves long-distance free-space teleportation
path between a satellite and a ground station. Here we report such of an independent quantum state, thus paving the way for satellite-
an experiment, using active feed-forward in real time. The experi- based global quantum communication.
ment uses two free-space optical links, quantum and classical, over Quantum teleportation relies on using both a quantum channel and
143 kilometres between the two Canary Islands of La Palma and a classical channel between two parties, usually called Alice and Bob5
Tenerife. To achieve this, we combine advanced techniques invol- (here located in La Palma and Tenerife respectively; Fig. 1a). The
ving a frequency-uncorrelated polarization-entangled photon pair quantum channel is used by Alice and Bob to share the entangled
source, ultra-low-noise single-photon detectors and entanglement- auxiliary state23
assisted clock synchronization. The average teleported state fidelity 1  
jY { i23 ~ pffiffiffi jH i2 jV i3 {jV i2 jH i3 ð1Þ
is well beyond the classical limit9 of two-thirds. Furthermore, 2
we confirm the quality of the quantum teleportation procedure which
  is one of the four maximally   entangled Bell states
without feed-forward by complete quantum process tomography. (Y + ~ p1ffiffi2 ðjH ijV i+jV ijH iÞ and W+ ~ p1ffiffi2 ðjH ijH i+jV ijV iÞ).
Our experiment verifies the maturity and applicability of such
jHæ and jVæ denote the horizontal and vertical polarization states. Alice
technologies in real-world scenarios, in particular for future
and Bob share this entangled state, where photon 2 is with Alice and
satellite-based quantum teleportation.
photon 3 is with Bob. Charlie provides the input photon 1 to be tele-
Significant progress has been made recently in the field of quantum ported to Alice in a general polarization state
communication based on optical free-space links10–19, which poten-
tially allow much larger propagation distances compared to the exist- jwi1 ~ajH i1 zbjV i1 ð2Þ
ing fibre networks because of the lower photon loss per kilometre. To where a and b are complex numbers (jaj2 1 jbj2 5 1), unknown to
enable quantum communication on a global scale and among parties both Alice and Bob.
not having access to any fibre network, it is foreseeable that experi- Alice then performs a Bell-state measurement (BSM), projecting
ments will involve ground-to-satellite10,11,15 and inter-satellite links. photons 1 and 2 randomly onto one of the four Bell states each with
Previous experiments focused on the distribution of quantum states the same probability of 25%. As a consequence, photon 3 is projected
of single photons or entangled photon pairs over optical free-space onto the input state jwæ, up to a unitary transformation (U), which
links. However, the realization of more sophisticated multiphoton depends on the outcome of the BSM. When Alice feeds the outcome of
quantum information protocols, such as quantum teleportation, the BSM forward to Bob via the classical channel, he can implement
remained an experimental challenge under real-world conditions. the corresponding unitary operation in real time and thus obtain
Quantum teleportation is based on the simultaneous creation of at photon 3 in the initial state (equation (2)) of photon 1. If jY2æ12 is
least three photons, which for random photon pair sources reduces detected, then U corresponds to the identity operation, which means
the count rate by several orders of magnitude compared to experi- that Bob needs to do nothing. If, on the other hand, jY1æ12 is detected,
ments using only two photons16,19. This decreases the signal-to-noise Bob has to apply a p phase shift between the horizontal and the vertical
ratio and requires a long integration time, such that high system component of his photon 3.
1
Institute for Quantum Optics and Quantum Information (IQOQI), Austrian Academy of Sciences, Boltzmanngasse 3, A-1090 Vienna, Austria. 2Vienna Center for Quantum Science and Technology, Faculty
of Physics, University of Vienna, Boltzmanngasse 5, A-1090 Vienna, Austria. 3Max Planck Institute of Quantum Optics, Hans-Kopfermann-Straße 1, 85748 Garching/Munich, Germany. 4Institute for
Quantum Computing and Department of Physics and Astronomy, University of Waterloo, 200 University Avenue West, Waterloo, Ontario N2L 3G1, Canada. {Present address: Department of Electrical
Engineering, Yale University, New Haven, Connecticut 06520, USA.

0 0 M O N T H 2 0 1 2 | VO L 0 0 0 | N AT U R E | 1
©2012 Macmillan Publishers Limited. All rights reserved
 RESEARCH LETTER

a La Palma Tenerife

φ3

Ψ 12– / Ψ 12+ Classical feed-forward channel

I/π
N
143 km
φ1 BSM
Quantum channel
0 1 2 3

HSP EPR
Ψ − 〉23

Bob
532 nm

Tx
b CCD Tracking
532 nm GPS
3 808 nm
EPR/Alice

TTU/Logic
Quantum channel
100 m

Rx
f
BBO1 Classical feed-
t
forward channel
BBO2 0 e

CCD
404 nm 143 km
fs pulsed Ψ+
80 MHz 2 1
HSP/Charlie
BSM/Alice

Ψ+
d
1,064 nm
c
TTU/Logic

Pump BBO Fibre Laser

Beam paths
Mirror Coil PD
a PBS
HWP Cable
QWP Coupler DM
b c APDc
IF 8 nm
FPC 7 cm diam. lens
IF 3 nm
GPS EOM
Ψ– Beam dump FBS 14 cm diam. lens

Figure 1 | Quantum teleportation between the Canary Islands La Palma and photon pair source generated photons 2 and 3 in BBO1 (EPR/Alice) and a
Tenerife over both quantum and classical 143-km free-space channels. collinear photon pair source generated photons 0 and 1 in BBO2 (HSP/Charlie).
a, Experimental scheme. Alice and Charlie are situated in La Palma, and Bob in All single photons were coupled into single-mode fibres. For implementing the
Tenerife. Charlie prepares the teleportation input photon 1 in | wæ1, using a BSM, photons 1 and 2 interfered in a fibre beam splitter (FBS) followed by
heralded single-photon (HSP) source with a trigger photon 0 (photons are polarization-resolving single-photon detection (BSM/Alice). Photon 3 was
indicated by black numerals on red circles). An Einstein–Podolsky–Rosen guided to the transmitter telescope via a 100-m single-mode fibre and sent to
(EPR)23 source generates an entangled pair of photons 2 and 3 in the state Bob in Tenerife, where the unitary transformation was implemented using an
| Y2æ23. Alice then performs a Bell-state measurement (BSM) on photons 1 and electro-optical modulator (EOM) and photon 3’s polarization was measured. A
2, and projects them onto two of the four Bell states ( | Y2æ12/ | Y1æ12), and sends real-time feed-forward operation was implemented by encoding the | Y1æ12
the result via the classical feed-forward channel to Bob. Photon 3 is sent via the BSM result in 1,064-nm laser pulses, which were then sent to Bob via the feed-
free-space quantum channel to Bob, who applies a unitary transformation forward channel. On Bob’s side, they were separated by a dichroic mirror (DM),
(identity operation or p phase shift) on photon 3 depending on the BSM result detected with a photodetector (PD) and used to trigger the EOM to perform the
and thus turns its state | wæ3 into a replica of the initial quantum state | wæ1. required p phase shift operation. See main text for details.
b, Set-up. In La Palma, a frequency-uncorrelated polarization-entangled

Our experiment was conducted between Alice’s transmitter station single-photon (HSP) source providing Charlie’s photon 1 to be tele-
(at the Jacobus Kapteyn Telescope (JKT) of the Isaac Newton Group ported. That source delivered pairs of horizontally (photon 0) and
on La Palma) and Bob’s receiver station (at the Optical Ground Station vertically (photon 1) polarized photons in a product state. The detec-
(OGS) of the European Space Agency on Tenerife), separated by tion of photon 0 by the avalanche photodiode (APD) t served as a
143 km, both at altitudes of about 2,400 m. Our experimental set-up trigger to herald the presence of photon 1. All photons were spectrally
is shown in Fig. 1b. At La Palma, near-infrared femtosecond pulses filtered using interference filters (IF) and coupled into single-mode
(with a central wavelength of 808 nm) emitted from a mode-locked fibres for spectral and spatial mode selection.
Ti:sapphire laser, with a repetition rate of 80 MHz, were up-converted To realize the BSM, Alice’s photon 2 was overlapped on a fibre beam
to blue pulses (with a central wavelength of 404 nm). They were used splitter (FBS) with the teleportation input photon 1, whose polariza-
to generate two photon pairs via type-II spontaneous parametric tion was arbitrarily prepared by Charlie using half- and quarter-wave
down conversion (SPDC) in two nonlinear b-barium borate (BBO) plates (HWP and QWP). In each output port of the FBS, a polarizing
crystals placed in sequence. The first SPDC source was aligned to beam splitter (PBS) was used to project these photons on either hori-
emit the entangled auxiliary photon pairs (photons 2 and 3) in the zontal or vertical polarization. Fibre polarization controllers (FPCs)
jY2æ23 state24, equation (1), while the second crystal was a heralded were used to compensate for unwanted polarization rotation induced
2 | N AT U R E | VO L 0 0 0 | 0 0 M O N T H 2 0 1 2
©2012 Macmillan Publishers Limited. All rights reserved
 LETTER RESEARCH

by the fibres. Our BSM can identify two out of the four Bell states, The real-life long-distance environment provided a number of
which is the optimum achievable with linear optics only25. For more challenges for the present teleportation experiment. These challenges
details on our BSM, see Supplementary Information. While the BSM resulted most significantly in the need to cope with an extremely low
was being performed at Alice, photon 3 was guided to a 7-cm-aperture signal-to-noise ratio when using standard techniques—indeed, too low
transmitter telescope through a 100-m-long single-mode fibre and to perform a successful experiment. To enhance the signal-to-noise
then sent via a 143-km free-space quantum channel over to Bob in ratio to a level that made the experiment possible, we used the follow-
Tenerife. There, it was collected by the 1-m-aperture OGS telescope, ing advanced techniques: a frequency-uncorrelated polarization-
and guided through its Coudé path to Bob. entangled photon pair source26–28, ultra-low-noise large-active-area
In the first stage of our experiment, we only considered the cases single-photon detectors at Bob29, and entanglement-assisted clock
where Alice detected jY2æ12 in the BSM, which results in photon 3 synchronization13,14,17. Whereas all these techniques have been imple-
being already in the state of the input photon, jwæ1, and hence Bob was mented individually, our work is (to our knowledge) the first that
required to perform an identity operation, that is, to do nothing at all. combines all of them simultaneously and moreover in an outdoor
We verified the success of the teleportation process by analysing the environment. See Supplementary Information for details.
polarization state of photon 3, which was accomplished by a polariza- First we present our results without feed-forward, where we
tion analyser, consisting of a quarter- and a half-wave plate and two only considered the BSM outcome jY2æ12. The input state jwæ1
free-space coupled Si-APDs (silicon avalanche photodiodes) placed in was always approximately one ofpffiffitheffi four ideal input
pffiffiffi states
each output mode of a polarizing beam splitter. jwideal i[ jH i,jV i,jPi~ðjH izjV iÞ 2,jLi~ðjH i{ijV iÞ 2 . We
In the second stage of our experiment, we implemented a real-time performed tomographic measurements on three consecutive nights,
feed-forward operation. When Alice obtained jY1æ12, she sent this thereby accumulating data over 6.5 h. Figure 2 shows the state tomo-
classical information to Bob. On receiving this information, Bob had graphy results of quantum teleportation. The measured density matrix r
to apply a p phase shift between the jHæ and jVæ components of photon for each of these teleported states was reconstructed from the experi-
3 to obtain the replica of the input state jwæ1. For further details on the mentally obtained data using the maximum-likelihood technique30. The
feed-forward implementation, see Supplementary Information. fidelity of the teleported state is defined as the overlap of the ideal tele-
The relevant events on Alice’s and Bob’s sides were recorded with ported state jwidealæ with the measured density matrix: f 5 Æwidealjrjwidealæ.
separate time-tagging units each disciplined to the global positioning For this set of states, the teleported state fidelities are measured to be
system (GPS)14. First, Alice in La Palma identified the threefold coin- f 5 0.890(42), 0.865(46), 0.845(27) and 0.852(37), yielding an average
cidence events corresponding to the jY6æ12 outcomes of the BSM. This f ~0:863ð38Þ, where digits in parentheses represent 1s uncertainties,
was done using a coincidence logic circuit featuring two separate out- for example, f ~0:863+0:038. During these measurements the link
put signals realized with transistor-transistor logic (TTL) pulses. These attenuation varied from 28.1 dB to 39.0 dB, which was mainly caused
TTL pulses were fed into a time-tagging unit which recorded the exact by rapid temperature change and strong wind. Despite such high loss in
time and the BSM result into a binary file. Similarly, Bob fed either the the quantum free-space channel, the classical average fidelity limit9 of
signals of both detectors (stage 1, without feed-forward) or the coin- 2/3 was clearly surpassed by our observed fidelities, as shown in Fig. 3.
cidence between these signals and the BSM results sent via 1,064-nm (Note that our random sampling of the input states over the mutually
laser pulses (stage 2, with feed-forward) into his time-tagging unit. unbiased bases states leads to the same classical limit as sampling over
After a measurement run was completed, both time-tagged data files the whole Bloch sphere. Also, it does not matter whether the sampling is
were compared by cross-correlation, and the detection events asso- done over all six mutually unbiased basis states or over three states with
ciated with simultaneous detection of four photons originating from one state per basis.) Therefore, we have explicitly demonstrated
the same pump pulse were identified. quantum teleportation over the 143-km free-space channel.

a b
1 1 1 1
Re( ρ)

Im( ρ)

Im( ρ)
Re( ρ)

0 0 0 0

–1 –1 –1 –1
H H H H H H H H
V V V V V V V V
ρideal = |H〉〈H| ρideal = |V〉〈V|

c d
1 1 1 1
Re( ρ)

Re( ρ)
Im( ρ)

Im( ρ)

0 0 0 0

–1 –1 –1 –1
H H H H H H H H
V V V V V V V V
1 1
ρideal = (|H〉〈H| + |H〉〈V| + |V〉〈H| + |V〉〈V|) ρideal = (|H〉〈H| + i|H〉〈V| – i|V〉〈H| + |V〉〈V|)
2 2

Figure 2 | State tomography results of the four quantum states without space channel. The wire grid indicates the expected values for the ideal cases.
feed-forward over the 143-km free-space channel with the BSM outcome of The data shown comprise a total of 605 fourfold coincidence counts in about
| Y2æ12. The bar graphs show the reconstructed density matrices r for the four 6.5 h. The uncertainties in state fidelities extracted from these density matrices
states teleported from Alice (La Palma) to Bob (Tenerife) over the 143-km free- are calculated using a Monte Carlo routine assuming Poissonian errors.

0 0 M O N T H 2 0 1 2 | VO L 0 0 0 | N AT U R E | 3
©2012 Macmillan Publishers Limited. All rights reserved
 RESEARCH LETTER

1.0 fidelity of 0.5, which is the limit one can reach with a classical strategy
where Alice and Bob do not share any entanglement as a resource.
In the second stage of the experiment, we realized quantum teleporta-
0.8 tion including real-time feed-forward of the BSM result over the143-km
pffiffiffi
classical channel. We set jPæ and jRæ states (jRi~ðjH izijV iÞ 2) as
input states for which the required p phase shift between the jHæ and the
Classical limit: 66.7%
jVæ components of photon 3 resulted in a 90u polarization rotation.
State fidelity

0.6
However, for the jHæ or jVæ input state, feed-forward is irrelevant
because a p phase shift would only result in a non-detectable global
0.4 phase shift. Thus, the quality of teleportation of these states is already
confirmed by our first-stage experiment. Realizing teleportation for the
states jPæ and jRæ fully confirms the generality of the procedure, as these
0.2 states belong to different mutually unbiased bases. In Tenerife, we
analysed photon 3 in the eigenbasis of the input state, that is, the jPæ/
jMæ (jRæ/jLæ) basis pffiffiwhen
ffi the input state was jPæ (jRæ). Here
0.0 jM i~ðjH i{jV iÞ 2. The resultant fidelities of the teleported states
|H〉 |V〉 |P〉 |L〉 |P〉 FF |R〉 FF are 0.760(50) and 0.800(37) for jPæ and jRæ, respectively (red bars in
Input states Fig. 3). Both results are clearly above the classical fidelity bound. Note
that in our experiment, the efficiency of the classical link was measured
Figure 3 | Summary of the state fidelity results for the teleported quantum
to be 21.3%, which was mainly due to amplitude fluctuations caused by
states with and without feed-forward. Grey bars show the results obtained in
quantum teleportation without feed-forward. Bob was informed via an internet atmospheric turbulence.
connection when Alice’s BSM outcome was | Y2æ12, that is, those cases where Using the real-time feed-forward operation, we unambiguously
he is not required to apply any operation to his photon. The uncertainties of experimentally demonstrated quantum teleportation from La Palma
these state fidelities are the same as presented in the main text. Red bars stand to Tenerife over a 143-km free-space channel. We note that from a
for the results obtained in quantum teleportation with feed-forward (FF). Bob conceptual perspective, real-time feed-forward is part of the original
was informed via the classical free-space link of Alice’s BSM outcomes of teleportation proposal5. Ultimately, the advantage of long-distance
| Y1æ12. The uncertainties of these state fidelities are derived from Poissonian teleportation compared to just sending the quantum state itself may
statistics. All observed fidelities significantly exceed the classical average fidelity
lie in the following future applications: if Alice and Bob can stockpile
limit of 2/3 (66.7%). Error bars, 61s.
their entangled states beforehand (with the help of quantum memories),
The reconstructed density matrices of the teleported quantum states teleportation is advantageous if the quantum channel is of low quality or
allow us to fully characterize the teleportation procedure by quantum if Bob’s location is unknown to Alice. This is because Alice can broadcast
process tomography. The four input states (rideal 5 jwidealæÆwidealj 5 the classical information with high quality and to arbitrary, or even
jHæ ÆHj, jVæ ÆVj, jPæ ÆPj, jLæ ÆLj) are transferred to the corresponding unknown, locations5. (Also, the quantum repeater, which is of high
(reconstructed) output states r. We can completely describe the effect of importance for large-scale quantum networks, is based on teleportation
teleportation on the inputPstates rideal by determining the process in the form of entanglement swapping31.) Specifically, this advantage
matrix x, defined by r~ 3l,k~0 xlk sl rideal sk , where the si are the also shows the usefulness of quantum teleportation even for quantum
Pauli matrices with s0 the identity operator. The process matrix x channels that allow only small transmission rates. We note that over the
can be computed analytically from these four equations for the four years teleportation sources have been markedly improved. We believe
different input and output states2. The ideal process matrix of quantum this development will continue and currently several new schemes are
teleportation xideal has only one non-zero component, (xideal)00 5 1, being pursued.
meaning the input state is teleported without any reduction in quality. Our work proves the feasibility of both ground-based and satellite-
Figure 4a and b show respectively the real and imaginary components of based free-space quantum teleportation. Our quantum teleportation
x for quantum teleportation based on our experimental results. The set-up was able to achieve coincidence production rates and fidelities
process fidelity of our experiment was fprocess 5 Tr (xidealx) 5 0.710(42). sufficient to cope with the optical link attenuation resulting from
This clearly confirmed the quantum nature of our teleportation experi- various experimental and technical challenges; such challenges will
ment as it is five standard deviations above the maximum process arise in quantum transmission between a ground-based transmitter

a b
1 1

0.8 0.8

0.6 0.6
Re(χ)

Im(χ)

0.4 0.4

0.2 0.2

0 0
σ0 σ0 σ0 σ0
σ1 σ1 σ1 σ1
σ2 σ2 σ2 σ2
σ3 σ3 σ3 σ3

Figure 4 | Quantum process tomography of quantum teleportation without teleportation. The operators si are the identity (i 5 0) and the x-, y-, and z-Pauli
feed-forward. a, b, The real (Re(xlk); a) and imaginary (Im(xlk); b) values of the matrices (i 5 1, 2, 3). For the ideal case, the only non-zero component of the
components of the reconstructed quantum process matrix, with l, k 5 0, 1, 2, process matrix of quantum teleportation, xideal, is (xideal)00 5 1, which is
and 3. The results of the state tomography of the four teleported states, | Hæ, indicated by the wire grid. Our experiment clearly confirmed that x00 (identity
| Væ, | Pæ, | Læ, are employed to reconstruct the process matrix of quantum state operation) is indeed the dominant component and surpasses the classical limit.

4 | N AT U R E | VO L 0 0 0 | 0 0 M O N T H 2 0 1 2
©2012 Macmillan Publishers Limited. All rights reserved
 LETTER RESEARCH

and a low-Earth-orbiting satellite receiver32. In fact, some of the 21. Ursin, R. et al. Quantum teleportation across the Danube. Nature 430, 849 (2004).
22. Briegel, H.-J., Dür, W., Cirac, J. I. & Zoller, P. Quantum repeaters: the role of
demands of our experiment were even more challenging than in imperfect local operations in quantum communication. Phys. Rev. Lett. 81,
satellite communication since there the atmospheric distances to be 5932–5935 (1998).
overcome are certainly shorter than the distance between Tenerife and 23. Einstein, A., Podolsky, B. & Rosen, N. Can quantum-mechanical description of
La Palma. In addition, acquiring, pointing and tracking (APT) for physical reality be considered complete? Phys. Rev. 47, 777–780 (1935).
24. Kwiat, P. G. et al. New high-intensity source of polarization-entangled photon pairs.
diffraction-limited transmitter and receiver telescopes in space is a Phys. Rev. Lett. 75, 4337–4341 (1995).
well-established technology. Therefore, our experiment represents a 25. Calsamiglia, J. & Lütkenhaus, N. Maximum efficiency of a linear-optical Bell-state
crucial step towards future quantum networks in space, which require analyzer. Appl. Phys. B 72, 67–71 (2001).
26. Kim, Y.-H., Kulik, S. P., Chekhova, M. V., Grice, W. P. & Shih, Y. Experimental
space-to-ground quantum communication. During the finalization of entanglement concentration and universal Bell-state synthesizer. Phys. Rev. A 67,
our manuscript, a related work was reported33 that used a 150-Hz APT 010301(R) (2003).
system, while our APT system operated at a few hertz, yet fast enough to 27. Poh, H. S., Lim, J., Marcikic, I., Lamas-Linares, A. & Kurtsiefer, C. Eliminating spectral
fulfil the specific requirements of our experiment. The most significant distinguishability in ultrafast spontaneous parametric down-conversion. Phys.
Rev. A 80, 043815 (2009).
distinction between ref. 33 and the experiment presented in this Letter 28. Yao, X.-C. et al. Observation of eight-photon entanglement. Nature Photon. 6,
is our implementation of an active feed-forward technique. The 225–228 (2012).
technology implemented in both experiments has certainly reached 29. Kim, Y.-S., Jeong, Y.-C., Sauge, S., Makarov, V. & Kim, Y.-H. Ultra-low noise single-
the required maturity for both satellite and long-distance ground com- photon detector based on Si avalanche photodiode. Rev. Sci. Instrum. 82, 093110
(2011).
munication. We expect that many of the features implemented here will 30. White, A. G., James, D. F. V., Eberhard, P. H. & Kwiat, P. G. Nonmaximally entangled
be important building blocks for a new area of experiments. states: production, characterization, and utilization. Phys. Rev. Lett. 83,
3103–3107 (1999).
Received 21 May; accepted 14 August 2012. 31. Żukowski, M., Zeilinger, A., Horne, M. A. & Ekert, A. K. ‘‘Event-ready-detectors’’
Bell experiment via entanglement swapping. Phys. Rev. Lett. 71, 4287–4290
Published online 5 September 2012.
(1993).
32. Aspelmeyer, M., Jennewein, T., Pfennigbauer, M., Leeb, W. R. & Zeilinger, A. Long-
1. Kimble, H. J. The quantum internet. Nature 453, 1023–1030 (2008).
2. Nielsen, M. & Chuang, I. Quantum Computation and Quantum Information distance quantum communication with entangled photons using satellites. IEEE J.
(Cambridge Univ. Press, 2000). Sel. Top. Quantum Electron. 9, 1541–1551 (2003).
3. Gisin, N. & Thew, R. Quantum communication. Nature Photon. 1, 165–171 (2007). 33. Yin, J. et al. Quantum teleportation and entanglement distribution over 100-
4. Ladd, T. D. et al. Quantum computers. Nature 464, 45–53 (2010). kilometre free-space channels. Nature 488, 185–188 (2012).
5. Bennett, C. H. et al. Teleporting an unknown quantum state via dual classical and Supplementary Information is available in the online version of the paper.
Einstein-Podolsky-Rosen channels. Phys. Rev. Lett. 70, 1895–1899 (1993).
6. Bouwmeester, D. et al. Experimental quantum teleportation. Nature 390, 575–579 Acknowledgements We thank the staff of IAC: F. Sanchez-Martinez, A. Alonso,
(1997). C. Warden, M. Serra and J. Carlos; and the staff of ING: M. Balcells, C. Benn, J. Rey,
7. Boschi, D., Branca, S., De Martini, F., Hardy, L. & Popescu, S. Experimental O. Vaduvescu, A. Chopping, D. González, S. Rodrı́guez, M. Abreu, L. González; J. Kuusela,
realization of teleporting an unknown pure quantum state via dual classical and E. Wille and Z. Sodnik; and J. Perdigues of the OGS and ESA. X.-S.M., T.J., R.U. and A.Z.
Einstein-Podolsky-Rosen channels. Phys. Rev. Lett. 80, 1121–1125 (1998). thank S. Ramelow for discussions, P. Kolenderski for discussions on the SPDC source
8. Furusawa, A. et al. Unconditional quantum teleportation. Science 282, 706–709 with the Bell-state synthesizer, S. Zotter for help during the early stages of the
(1998). experiment, and R. Steinacker for meteorological advice. J.K. was supported by the EU
9. Massar, S. & Popescu, S. Optimal extraction of information from finite quantum project MALICIA. E.A. and V.M. thank C. Kurtsiefer and Y.-S. Kim for detector electronics
ensembles. Phys. Rev. Lett. 74, 1259–1263 (1995). design, J. Skaar for support, and the Research Council of Norway (grant No. 180439/
10. Hughes, R. J. et al. Free-space quantum key distribution in daylight. J. Mod. Opt. 47, V30) and Industry Canada for support. This work was made possible by grants from the
549–562 (2000). European Space Agency (contract 4000104180/11/NL/AF), the Austrian Science
11. Rarity, J. G., Tapster, P. R., Gorman P. M. & Knight, P. Ground to satellite secure key Foundation (FWF) under projects SFB F4008 and CoQuS, and the FFG for the QTS
exchange using quantum cryptography. N. J. Phys. 4, 82 (2002). project (no. 828316) within the ASAP 7 program. We also acknowledge support by the
12. Aspelmeyer, M. et al. Long-distance free-space distribution of quantum European Commission, grant Q-ESSENCE (no. 248095) and the John Templeton
entanglement. Science 301, 621–623 (2003). Foundation.
13. Marcikic, I., Lamas-Linares, A. & Kurtsiefer, C. Free-space quantum key distribution
with entangled photons. Appl. Phys. Lett. 89, 101122 (2006). Author Contributions X.-S.M. conceived the research, designed and carried out the
14. Ursin, R. et al. Entanglement-based quantum communication over 144 km. Nature experiment, and analysed data. T.H., T.S. and D.W. carried out the experiment and
Phys. 3, 481–486 (2007). analysed data. S.K., W.N., B.W. and A.M. provided experimental assistance during the
15. Villoresi, P. et al. Experimental verification of the feasibility of a quantum channel early stage of the experiment. J.K. provided the theoretical analysis and analysed data.
between space and Earth. N. J. Phys. 10, 033038 (2008). E.A. and V.M. developed the ultra-low-noise detectors. T.J. provided experimental and
16. Fedrizzi, A. et al. High-fidelity transmission of entanglement over a high-loss free- conceptual assistance, and conceived and developed the coincidence analysis code.
space channel. Nature Phys. 5, 389–392 (2009). R.U. conceived the research, planned and carried out the experiment and analysed
17. Scheidl, T. et al. Feasibility of 300 km quantum key distribution with entangled data. A.Z. defined the scientific goals, conceived the research, designed the experiment
states. N. J. Phys. 11, 085002 (2009). and supervised the project. X.-S.M., T.H., T.S., J.K., R.U. and A.Z. wrote the manuscript
18. Jin, X.-M. et al. Experimental free-space quantum teleportation. Nature Photon. 4, with assistance from all other co-authors.
376–381 (2010).
19. Scheidl, T. et al. Violation of local realism with freedom of choice. Proc. Natl Acad. Author Information Reprints and permissions information is available at
Sci. USA 107, 19708–19713 (2010). www.nature.com/reprints. The authors declare no competing financial interests.
20. Marcikic, I., de Riedmatten, H., Tittel, W., Zbinden, H. & Gisin, N. Long-distance Readers are welcome to comment on the online version of the paper. Correspondence
teleportation of qubits at telecommunication wavelengths. Nature 421, 509–513 and requests for materials should be addressed to X.-S.M. ([email protected])
(2003). or A.Z. ([email protected]).

0 0 M O N T H 2 0 1 2 | VO L 0 0 0 | N AT U R E | 5
©2012 Macmillan Publishers Limited. All rights reserved