Attractive Photons in A Quantum Nonlinear Medium: Citation
Attractive Photons in A Quantum Nonlinear Medium: Citation
Attractive Photons in A Quantum Nonlinear Medium: Citation
The MIT Faculty has made this article openly available. Please share
how this access benefits you. Your story matters.
Citation
As Published
http://dx.doi.org/10.1038/nature12512
Publisher
Version
Accessed
Citable Link
http://hdl.handle.net/1721.1/91605
Terms of Use
Detailed Terms
this work.
Our experiment (outlined in Fig. 1a) makes use of an
ultracold rubidium gas loaded into a dipole trap, as described previously19 . The probe light of interest is +
polarized, coupling the ground state |gi to the Rydberg state |ri via an intermediate state |ei of linewidth
/(2) = 6.1 MHz by means of a control field that is detuned by below the resonance frequency of the upper
transition |ei |ri (Fig. 1b). Under these conditions, for
a very weak probe field with mean incident photon rate
Ri = 0.5 s1 , EIT is established when the probe detuning matches that of the control field (see Fig. 1c showing
the probe transmission and phase shift). However, the
Rydberg medium is extremely nonlinear: a probe photon rate of Ri = 5 s1 saturates the medium due to
the Rydberg blockade25 , yielding a probe spectrum close
to the bare two-level response. Given the measured system bandwidth of about 5 s1 , this implies a substantial nonlinear response with average pulse energies corresponding to less than one photon per inverse bandwidth.
We perform our experiments on two-photon resonance
|gi |ri, where the transmission is approximately independent of the probe photon rate for our experimental
parameters at || > , yielding a purely dispersive nonlinearity. The linear dispersion at this point corresponds
to a reduced probe group velocity of typically vg = 400
m/s, while the group velocity dispersion endows the photons with an effective mass26 m 1000~/c2 , where
is the optical frequency and c is the speed of light in
vacuum.
In order to explore the quantum dynamics in the propagation of photon pairs, we measure time-dependent twophoton correlation functions of the transmitted light (see
Fig. 1a). To determine both amplitude and phase of the
+ -polarized probe field, we prepare input coherent light
| 1001/2
Control
= + +
| 53/2
PBS
d
2
++
1
0.5
0
-20
Phase [rad]
( )
20
2
0
1
Probe ( + )
| 51/2
1.5
1
0.5
-20
20
-20
0
r/r
/
B
20
Transmission
1/15
1
1
0
-1
-2
-20
0
20
Probe frequency detuning [(2) MHz]
FIG. 1: Photons exhibiting strong mutual attraction in a quantum nonlinear medium. a,b, A linearly polarized
weak laser beam near the transition |gi |ei at 780 nm is sent into a cold rubidium gas driven by a control laser near the
transition |ei |ri at 479 nm. Strong nonlinear interactions between + -polarized photons are detected via photon-photon
correlation functions of the transmitted light for a set of different polarization bases, as determined by a quarter-waveplate
(QWP), a half-waveplate (HWP), and a polarizing beam-splitter (PBS). Here photons serve as a phase reference. c,
Transmission spectra (top) and phase shift (bottom) for + photons with incoming rate of Ri = 0.5 s1 (blue squares) and
Ri = 5 s1 (green circles), for a control field red-detuned by /(2) = 15 MHz (blue line is theory). The spectrum at high
probe rate approaches that of the undriven two-level system (dashed gray; see also Fig. SI2). The solid vertical line corresponds
to the EIT resonance. d, Photon bunching and two-photon bound state. Theoretically predicted photon-photon correlation
function in the Schroedinger-equation approximation (top, blue line) for /(2) = 14 MHz, with a potential well of width 2rB
(bottom, green line). The bound state (bottom, red) and the superposition of scattering states (bottom, black) form the initial
wave function = 1 (bottom, dashed blue). The two-photon bound state results in the observed bunching in the correlation
(2)
function g++ ||2 (top, gray circles), where time has been converted into distance via the group velocity vg . The boundary
effects resulting from the finite extent of the atom cloud become important for |r| 5 rB .
The measured g allow us to reconstruct the twophoton density matrix using quantum state tomography via a maximum-likelihood estimation27,28 , from
which we define an interaction matrix by factoring out
the linear response, such that directly quantifies the
nonlinearity (see Methods). The density matrix approach
is necessary to account for decoherence and technical imperfections. The probability density of two interacting
(2)
+ photons, g++ = ++,++ , and the nonlinear phase,
acquired by the + + pair relative to a non-interacting
pair, = arg [
++, ], are shown in Figs. 2a,b for
/(2) = 14 MHz. The time dependence allows us to
extract the nonlinear phase as a function of the photonphoton separation. Clearly visible is the bunching of pho-
++
1.5
1.2
1
1
0.8
0.5
1.5
0
0
1 (s)
2
1.5
2 (s)
(rad)
0.2
0
-0.2
0.5
(rad)
1
0.5
0.5
0
0
1.5
2 (s)
++
0.2
0.4
[s]
(s)
0
-/8
-/4
-0.4
0
0
0.5
1.5
1 (s)
0.2
0.4
(s)
[s]
FIG. 2: Propagation of interacting photon pairs. Measured second-order correlation function (a) and nonlinear phase
shift (b) of interacting photon pairs at = 2.3. The photons are detected at times t1 and t2 . (c) Second-order correlation
function displayed as a function of the time difference | | = |t1 t2 | between the photons, showing the transition from antibunching on resonance ( = 0, green) to bunching at large detuning ( = 2.3, blue). Points are experimental data, lines are
(2)
full numerical simulations. All g++ measurements are rescaled by their value at > 1.5 s (See SI). (d) Nonlinear phase-shift
versus | | for two different detunings ( = 1.5, purple, and = 2.3, blue). The 1 error is 30 mrad, dominated by photon
shot noise.
medium can be understood by first considering an idealized situation with no decoherence between the Rydberg
state and the ground state. Then the steady-state in a
one-dimensional homogenous medium can be described
by a two-photon wavefunction (z1 , z2 ), whose evolution
is approximately governed by a simple equation19 in the
center-of-mass R = (z1 + z2 )/2 and relative r = z1 z2
coordinates:
2
2
V(r)
iR = 4la i +
V(r) 2c r2 +
.
(1)
la
6 6 1
Here the effective potential V(r) = [i+2
(1+2r /rB )]
1
approaches (i + 2/)
inside the blockaded volume
(|r| < rB ), and zero outside. The solution relates to
our measurements in the time domain forqsmall | | ap(2)
proximately via (R = L, r = vg )
g++ ( )ei( )
(see SI for the exact relation). Far off resonance (||
, c ), Eq. (1) corresponds to a Schroedinger equation
with the center-of-mass propagation distance R playing
the role of effective time. The photons effective mass
m /(16la ) can be positive or negative depending
on the sign of the detuning . As the sign of the potential also changes with (potential well for < 0, barrier for > 0), the effective force (F in Fig. 1a) in both
a
2
++ 0
2
1
0
-4
-2
-1
0
/
-3
-2
-1
0
/
/2
(0)
-3
-/2
-4
2.5
b vg [m]
c
2
++ 0
2
1.5
1
-3
-2 -1 0
/(2) [MHz]
60
40
20
0
-3
-2 -1 0
/(2) [MHz]
then be understood as arising from the interference between the bound and scattering states that evolve at different frequencies, and the observed bunching feature in
(2)
g++ reflects the wavefunction of the two-photon bound
state (see SI). As shown in Figs. 3a,b, the solution of
the Schroedinger-like equation (1) with a simplified deltafunction potential captures the essential features of the
nonlinear two-photon propagation. Additional experimental evidence for the bound-state dynamics is obtained
by tuning the probe field relative to the EIT resonance,
thereby varying the strength of the two-photon interaction potential. As the probe detuning approaches the
Raman resonance, the difference in refractive indices inside and outside the blockade radius increases and the
potential deepens (see SI and Fig. 1c). Consequently,
the bound state becomes more localized and the bunch(2)
ing, quantified by g++ (0), is enhanced, as evidenced in
Figs. 3c,d. Note that the size of the two-photon bound
state and correspondingly the width of the bunching feature 2b vg 70 m, exceed the width of the potential
well of 2rB 35 m, as expected for a potential with
one weakly bound state.
Figures 2 and 3 also show the results of our full theoretical model, in which we numerically solve the set of propagation equations for the light field and atomic coherences.
The model incorporates the longitudinal atomic-density
distribution and the decoherence of the Rydberg state
(see SI for details). These simulations are in good agreement with our experimental results and the predictions
of the simplified model, Eq. (1), confirming that the evolution of the two-photon wavepacket is dominated by the
attractive force between the photons.
Finally, we study the quantum coherence and polarization properties of the transmitted photon pairs. Figure
4a compares the purity of the two-photon density matrix
( ), that includes photon interactions, to the purity of
the product of one-photon matrices (1) (1) for noninteracting photons. At large photon separation , the
purity P ( ) of the two-photon density matrix is dominated by the one-photon decoherence due to partial depolarization of the transmitted light (see SI). This depolarization is attributed to the difference in group delay d be+
5
Methods
b
Concurrence C(||)
Concurrence
PurityP(||)
()
Purity
a 0.8
0.7
0.6
0
0.5
||
(s)
(s)
0.1
0.05
0
0
0.1
0.2
||
(s)
(s)
0.3
FIG. 4: Quantum coherence and entanglement. a, Purity P ( ) = Tr[( )2 ] of the measured two-photon densitymatrix for = 2.3 (blue symbols), approaching at large
photon separation the purity expected from the measured
2
one-photon density-matrix Tr[((1) (1) ) ] (dotted black
+ +
line). Interacting photon pairs near = 0 exhibit
lower decoherence. Error bars (1) are derived from the uncertainty in the density matrix due to detection shot noise. b,
Concurrence C( ) calculated from , indicating polarization
entanglement of proximal photons upon transmission through
the quantum nonlinear medium.
in the two-qubit basis 1+ 2+ , |Si , 1 2 , |Ai ,
where |S/Ai = (1+ 2 1 2+ )/ 2. Since the
two photons share the same frequency and spatial
mode, there is no coherence between the 3 3 symmetric and 1 1 anti-symmetric subspaces28 . We
measure in six required polarization bases, chosen as
6
its elements are equal to 1 in the absence of nonlinearity. Figure SI3 compares the measured photon-photon
correlation functions to those calculated from (see also
Figure SI4). The colormaps in Fig. 2 presenting values
derived from (t1 , t2 ) have been smoothed using an unweighted, nearest-neighbor, rectangular sliding-average.
Acknowledgements
We thank H. P. B
uchler, T. Pohl, J. Otterbach, P.
Strack, M. Gullans, and S. Choi for useful discussions.
10
11
12
13
14
15
16
Scully, M. O. & Zubairy, M. S. Quantum Optics (Cambridge University Press, Cambridge, UK, 1997).
Milburn, G. J. Quantum optical fredkin gate. Phys. Rev.
Lett. 62, 21242127 (1989).
Imamo
glu, A., Schmidt, H., Woods, G. & Deutsch, M.
Strongly interacting photons in a nonlinear cavity. Phys.
Rev. Lett. 79, 14671470 (1997).
Deutsch, I. H., Chiao, R. Y. & Garrison, J. C. Diphotons in a nonlinear fabry-perot resonator: Bound states of
interacting photons in an optical quantum wire. Phys.
Rev. Lett. 69, 36273630 (1992).
Shen, J.-T. & Fan, S. Strongly correlated two-photon
transport in a one-dimensional waveguide coupled to a twolevel system. Phys. Rev. Lett. 98, 153003 (2007).
Chang, D. E., Srensen, A. S., Demler, E. A. & Lukin,
M. D. A single-photon transistor using nanoscale surface
plasmons. Nat. Phys. 3, 807 812 (2007).
Cheng, Z. & Kurizki, G. Optical multiexcitons: Quantum gap solitons in nonlinear bragg reflectors. Phys. Rev.
Lett. 75, 34303433 (1995).
Turchette, Q. A., Hood, C., Lange, W., Mabuchi, H. &
Kimble, H. Measurement of conditional phase shifts for
quantum logic. Phys. Rev. Lett. 75, 4710 (1995).
Chang, D. E. et al. Crystallization of strongly interacting
photons in a nonlinear optical fibre. Nat. Phys. 4, 884889
(2008).
Fushman, I. et al. Controlled phase shifts with a single
quantum dot. Science 320, 769772 (2008).
Rauschenbeutel, A. et al. Coherent operation of a tunable
quantum phase gate in cavity qed. Phys. Rev. Lett. 83,
5166 (1999).
Kirchmair, G. et al. Observation of quantum state collapse
and revival due to the single-photon kerr effect. Nature
(London) 495, 205209 (2013).
Saffman, M., Walker, T. G. & Mlmer, K. Quantum information with rydberg atoms. Rev. Mod. Phys. 82, 2313
2363 (2010).
Gorshkov, A. V., Otterbach, J., Fleischhauer, M., Pohl, T.
& Lukin, M. D. Photon-photon interactions via rydberg
blockade. Phys. Rev. Lett. 107, 133602 (2011).
Pritchard, J. D. et al. Cooperative atom-light interaction
in a blockaded rydberg ensemble. Phys. Rev. Lett. 105,
193603 (2010).
Maxwell, D. et al. Storage and control of optical photons
using rydberg polaritons. Phys. Rev. Lett. 110, 103001
17
18
19
20
21
22
23
24
25
26
27
28
29
30
(2013).
Dudin, Y. O. & Kuzmich, A. Strongly interacting rydberg
excitations of a cold atomic gas. Science 336, 887889
(2012).
Petrosyan, D., Otterbach, J. & Fleischhauer, M. Electromagnetically induced transparency with rydberg atoms.
Phys. Rev. Lett. 107, 213601 (2011).
Peyronel, T. et al. Quantum nonlinear optics with single
photons enabled by strongly interacting atoms. Nature
(London) 488, 5760 (2012).
Parigi, V. et al.
Observation and measurement of
interaction-induced dispersive optical nonlinearities in an
ensemble of cold rydberg atoms. Phys. Rev. Lett. 109,
233602 (2012).
Kasapi, A., Jain, M., Yin, G. Y. & Harris, S. E. Electromagnetically induced transparency: Propagation dynamics. Phys. Rev. Lett. 74, 24472450 (1995).
Venkataraman, V., Saha, K. & Gaeta, A. L. Phase modulation at the few-photon level for weak-nonlinearity-based
quantum computing. Nature Photonics 7, 138141 (2012).
Rajapakse, R. M., Bragdon, T., Rey, A. M., Calarco, T.
& Yelin, S. F. Single-photon nonlinearities using arrays of
cold polar molecules. Phys. Rev. A 80, 013810 (2009).
Drummond, P. D. & He, H. Optical mesons. Phys. Rev.
A 56, R1107R1109 (1997).
Lukin, M. D. et al. Dipole blockade and quantum information processing in mesoscopic atomic ensembles. Phys.
Rev. Lett. 87, 037901 (2001).
Fleischhauer, M., Imamoglu, A. & Marangos, J. P. Electromagnetically induced transparency: Optics in coherent
media. Rev. Mod. Phys. 77, 633673 (2005).
James, D. F. V., Kwiat, P. G., Munro, W. J. & White,
A. G. Measurement of qubits. Phys. Rev. A 64, 052312
(2001).
Adamson, R. B. A., Shalm, L. K., Mitchell, M. W. & Steinberg, A. M. Multiparticle state tomography: Hidden differences. Phys. Rev. Lett. 98, 043601 (2007).
Sevincli, S., Henkel, N., Ates, C. & Pohl, T. Nonlocal
nonlinear optics in cold rydberg gases. Phys. Rev. Lett.
107, 153001 (2011).
Heidemann, R. et al. Evidence for coherent collective rydberg excitation in the strong blockade regime. Phys. Rev.
Lett. 99, 163601 (2007).