See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/237354782

Quantum interference by a nonlocal double slit

Article in Physical review A, Atomic, molecular, and optical physics · August 1999
DOI: 10.1103/PhysRevA.60.1530

CITATIONS READS
76 382

4 authors:

E. J. S. Fonseca Paulo H. Souto Ribeiro

Federal University of Alagoas Universidade Federal de Santa Catarina
117 PUBLICATIONS 2,204 CITATIONS 156 PUBLICATIONS 6,183 CITATIONS

SEE PROFILE SEE PROFILE

S. Pádua Carlos H Monken

Federal University of Minas Gerais Federal University of Minas Gerais
141 PUBLICATIONS 4,387 CITATIONS 131 PUBLICATIONS 4,717 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Carlos H Monken on 22 January 2014.

The user has requested enhancement of the downloaded file.

PHYSICAL REVIEW A VOLUME 60, NUMBER 2 AUGUST 1999

Quantum interference by a nonlocal double slit

E. J. S. Fonseca,1 P. H. Souto Ribeiro,2 S. Pádua,1 and C. H. Monken1,*
1
Departamento de Fı́sica, Universidade Federal de Minas Gerais, Caixa Postal 702, Belo Horizonte, MG 30123-970, Brazil
2
Instituto de Fı́sica, Universidade Federal do Rio de Janeiro, Caixa Postal 68.528, Rio de Janeiro 22945-970, RJ, Brazil
共Received 18 November 1998兲
We report an interference experiment in which photon pairs generated by spontaneous parametric down-
conversion produce a Young-type fourth-order interference pattern after being scattered by two different and
spatially separated apertures, whose superposition defines a double slit. The experiment is compared with
previous ones based on parametric down-conversion, and its nonlocal nature is discussed. A theoretical expla-
nation is also provided. 关S1050-2947共99兲06507-5兴

PACS number共s兲: 42.50.Ar, 03.65.Bz, 42.50.Dv

I. INTRODUCTION cal axis lies in the horizontal plane, pumped by an Argon

laser operating at 351.1 nm, TEM00 power stabilized mode,
Entanglement is one of the most striking features of quan- with an output power of approximately 50 mW. The crystal
tum mechanics. It is the basis of the exciting new fields of was cut for type II phase matching at ␭ 1 ⫽␭ 2 ⫽702 nm, with
quantum computing, quantum cryptography 关1兴, and telepor- output angles of approximately 4°, with respect to the pump-
tation 关2,3兴. Some nonlocal 关4兴 aspects of entanglement have ing direction. Detectors D 1 and D 2 were avalanche photo-
been explored in a variety of experimental situations in quan- diodes operating in the photon counting mode, placed at 100
tum optics, from which one can cite the Einstein-Podolsky- cm from the crystal. In front of each detector there was an
Rosen experiments 关5–10兴, two-photon optics 关11兴, trans- arrangement composed by a slit of 0.1 mm ⫻5 mm hori-
verse interference effects 关12–14兴, and conditional zontally aligned, followed by a microscope objective focused
interference 关15兴. In the latter three, the momentum entangle- on the active area. D 1 and D 2 were connected to single and
ment causes the fourth-order spatial correlation function of
coincidence counters with a resolving time of 10 ns. The
the electromagnetic field to be dependent on the relative
transversal position of two spatially separated photodetec- apertures A 1 and A 2 were chosen as indicated in Fig. 1,
tors. Recently, Monken, Souto Ribeiro, and Pádua 关14兴 where A 1 was a 0.4-mm⫻10-mm single slit, and A 2 was a
showed that the two-photon optics 关11兴 can be understood as 0.2-mm-diameter wire. Both A 1 and A 2 were placed at z A
a transfer of angular spectrum from the pump beam to the ⫽48 cm from the crystal. A convergent lens L of 50-cm
two-photon field generated by the process of spontaneous focal length was placed in the pump beam, so that the beam
parametric down-conversion in a thin crystal. The transfer of waist w 0 was located at 48 cm from the crystal, right on the
angular spectrum provides an easy way to control the fourth- plane of the apertures. The measured beam diameter at this
order spatial correlation of that field. position was 2w 0 ⫽0.15 mm.
In this paper we discuss an interesting consequence of Coincidence rates, which are proportional to the fourth-
momentum entanglement: the ability of a two-photon field to order correlation function, were recorded as functions of de-
mimic the scattering by a double slit, when this field is scat- tectors’ D 1 and D 2 vertical positions. Single count rates were
tered by two spatially separated apertures, none of them be- also recorded.
ing a double slit. Rather, the superposition of the two aper-
tures do define a double slit, which determines the shape of III. THEORY
the two-point fourth-order transverse spatial correlation func-
tion. The situation is represented in Fig. 1, where the en- In order to account for the effect of the apertures in the
tangled two-photon field is generated by spontaneous para- quantized field, let us start from the classical solution. Sup-
metric down-conversion 共SPDC兲.
An extension of the theory developed in Ref. 关14兴 is
shown to be able to describe our experimental results. We
also compare the experiment with previous ones
关7–10,12,13,16–20兴, emphasizing the nonlocal character of
the interference effects.

II. EXPERIMENT

The experimental setup is represented in Fig. 2. The

down-converter was a 7-mm-long BBO crystal, whose opti-

*Author to whom correspondence should be addressed. Electronic FIG. 1. Top: basic arrangement for the experiment. Bottom:
address: [email protected] detail of the apertures.

1050-2947/99/60共2兲/1530共4兲/$15.00 PRA 60 1530 ©1999 The American Physical Society

PRA 60 QUANTUM INTERFERENCE BY A NONLOCAL DOUBLE SLIT 1531

FIG. 2. Experimental setup.

pose that a monochromatic field E(r), propagating around FIG. 3. Coincidence counts when detector D1 is moved. Solid
the z direction, is known at the plane z⫽0. In terms of its line: numerical solution of Eq. 共9兲. Dotted line: solution of Eq. 共8兲.
spatial Fourier components, E(r) can be written as
(2,2)
共 r1 ,r2 兲 ⫽ 兩 g 共 r1 ,r2 兲 兩 2 , 共6兲
冕
G 12
iq• ␳
E 共 ␳,0 兲 ⫽ dqA共 q兲 e , 共1兲
where

where ␳ and q are the transverse components of r and k,

respectively. After being scattered by an aperture A( ␳)
g 共 r1 ,r2 兲 ⫽const 冕 ␰冕 d d ␩A 1 共 ␰兲 A 2 共 ␩兲 W
placed on the plane z⫽z A , the far field in the paraxial ap-
proximation is 关21兴
⫻ 冉 ␰ ␩
⫹ ,z A exp i 冊 冋 冉
k 兩 ␳1 ⫺ ␰兩 2
2 z 1 ⫺z A
冕 冕
2 2
E 共 ␳,z 兲 ⫽e ikz dq⬘ A共 q⬘ 兲 T 共 q⫺q⬘ 兲
冊册 冉 冊
dq
兩 ␳2 ⫺ ␩兩 2 兩 ␰⫺ ␩兩 2
⫹ exp ik , 共7兲

再冋 册冎
z 2 ⫺z A 4z A
q2 q ⬘2
⫻exp i q• ␳⫺ 共 z⫺z A 兲 ⫺ z , 共2兲
2k 2k A where W( ␳,z A ) is the transverse profile of the pump field at
z⫽z A . In order to simplify the calculations, the following
where T(q) is the Fourier transform of A( ␳). Now, replacing conditions were assumed: k 1 ⫽k 2 ⫽k, k 1 ⫹k 2 ⯝k p 共collinear
E(r) and A(q) by quantum mechanical operators, we arrive or quasicollinear phase matching兲, z A1 ⫽z A2 ⫽z A .
at the following operator for the scattered far field: Now, if we consider the particular case of a pump beam
focused at z⫽z A , so that W( ␳,z A ) can be approximated by a
E (⫹) 共 ␳,z 兲 ⫽e ikz 冕 冕
dq dq⬘ a 共 q⬘ 兲 T 共 q⫺q⬘ 兲
delta function, g(r1 ,r2 ) becomes

冕 冉 冊
␰2

再冋 册冎
g 共 r1 ,r2 兲 ⫽const. d ␰A 1 共 ␰兲 A 2 共 ⴚ␰兲 exp ik
q2 q ⬘2 4Z A
⫻exp i q• ␳⫺ 共 z⫺z A 兲 ⫺ . 共3兲

冋 册
z
2k A
冏 冏
2k
2
1
共 ␳ ⫺ ␳2 兲 ⫺ ␰
For a thin nonlinear crystal centered at the origin and 2 1
pumped along the z direction, the state generated by SPDC in ⫻exp ik . 共8兲
z d ⫺z A
the monochromatic and paraxial approximations can be rep-
resented by 关14兴 Again, for simplicity, we have considered z 1 ⫽z 2 ⫽z d . The
above expression describes the propagation of light from an
兩 ␺ 典 ⫽ 兩 v ac 典 ⫹const. 冕 冕
dq1 dq2 v共 q1 ⫹q2 兲 兩 1;q1 典 兩 1;q2 典 , effective aperture defined by A 1 ( ␰)A 2 (⫺ ␰) at z⫽z A to the
detection plane (z⫽z d ), in the paraxial approximation, with
共4兲 wave number 2k. When z d ⫺z A is large enough, this propa-
gation leads to the diffraction pattern of the effective aper-
where 兩 1;q典 represents a one-photon state with transverse ture as a function of the transverse coordinate 21 ( ␳1 ⫺ ␳2 ). In
wave vector component q, and v (q) is the angular spectrum practice, however, the dimensions of the apertures A 1 and A 2
of the pump field at z⫽0. are usually of the order of a few hundred ␮ m, so the focused
Combining expressions 共3兲 and 共4兲 we can calculate the pump beam profile can hardly be approximated by a delta
fourth-order correlation function function. A direct 共numerical兲 solution of expression 共7兲 is
then required, as discussed in Sec. IV.
(2,2)
G 12 共 r1 ,r2 兲 ⫽ 具 E (⫺)
2 共 ␳2 ,z 2 兲 E 1 共 ␳1 ,z 1 兲
(⫺)

IV. RESULTS AND DISCUSSION

1 共 ␳1 ,z 1 兲 E 2 共 ␳2 ,z 2 兲 典 ,
⫻E (⫹) 共5兲
(⫹)

Figure 3 shows the results of coincidence counts in sam-

when the apertures A 1 and A 2 are placed at z A1 and z A2 . pling times of 100 s, when detector D 1 is moved in steps of
After some algebra we have 0.15 mm, while detector D 2 is held at y 2 ⫽0. Figure 4 shows
1532 FONSECA, RIBEIRO, PÁDUA, AND MONKEN PRA 60

FIG. 4. Coincidence counts when detector D2 is moved. Solid FIG. 5. D1 (䊊) and D2 (䊉) single counts as functions of each
line: numerical solution of Eq. 共9兲. Dotted line: solution of Eq. 共8兲. detector position when the other is kept at zero.

the same kind of measurement when D 2 is moved while D 1 eter and the intensity pattern produced does not present
is held at y 1 ⫽0. In both cases, Young-type interference pat- fringes, we can regard this pattern as a continuous superpo-
terns are readily identified, although they do not refer to the sition of interference patterns displaced in space or time,
same effective aperture. This asymmetry is due to the finite resulting in fringes with extremely low visibility. When co-
diameter of the pump beam at z⫽z A , which is of the same incidence measurements are performed, a subensemble
order of magnitude as the dimensions of the apertures. A within that continuum of patterns is selected 关12兴.
numerical solution of expression 共6兲 in the one-dimensional In the present experiment with nonlocal double slits, this
case was carried out, that is, kind of local interpretation is not possible, since no Young

冏冕 冕
interference patterns can be associated with the intensity dis-
tributions in neither signal nor idler beams. It must be noted
(2,2)
G 12 共 y 1 ,y 2 ,z d 兲 ⫽const d␰ d ␩ A 1共 ␰ 兲 A 2共 ␩ 兲
that we are not claiming that our experiment shows nonlocal
interference effects and previous ones do not. The above

⫻exp ⫺ 冋冉 冊␰⫹␩
2w 0
2
⫹
ik
4z A
共 ␰⫺␩ 兲2 册 mentioned filtering process, due to coincidence detection be-
tween two spatially separated detectors, may exhibit a non-
local character. In those cases, however, the nonlocality can-

再冋 册冎冏
not be associated with the mere appearance of interference
共 y 1⫺ ␰ 兲2⫹共 y 2⫺ ␩ 兲2 2
⫻exp ik , fringes, but rather to their high visibility. In our case, it is
2 共 z d ⫺z A 兲 clear that the nonlocal character of the interference process is
explicit.
共9兲 We find an analogy between the present experiment and
two others. The first one was introduced by Rarity and Tap-
setting all parameters to the corresponding experimental val- ster 关20兴. In that case, it is possible to identify a nonlocal
ues. The solid lines in Figs. 3 and 4 are solutions of expres- Mach-Zehnder in which one of the beam splitters is placed in
sion 共9兲, normalized to the number of coincidences, with no the idler beam and the other one is placed in the signal. The
additional fitting, whereas the dotted lines represent the so- second experiment is the usual Bell state polarization inter-
lution of expression 共8兲, the limiting case. Figure 5 shows the ferometer 关6兴, where each of the polarization analyzers can
single counts of both detectors, which are expected from be viewed as part of a nonlocal interferometer in the spin
apertures A 1 and A 2 , illuminated by an extended source 共the space, since the coincidence pattern depends on the relative
interaction region in the crystal兲. angle between the analyzer’s axis. In both cases 关6,20兴, non-
The above results show that double-slit interference pat- locality is quantitatively evaluated by means of a Bell in-
terns are observed in the coincidence counting rate, even equality, which is violated.
though individual signal and idler intensity profiles have A Bell-type inequality for position variables was derived
nothing to do with double-slit patterns. Previous experiments by Ou 关22兴, but it does not apply to our experiment. Even
with double-slits and coincidence detection are based on the though we cannot evaluate it quantitatively, our experimental
passage of one of the beams 共signal 关12兴, idler 关13兴, or pump results suggest that strong nonlocal correlations are present.
关14兴兲 through a localized double slit. In other interference In conclusion, we have presented a quantum interference
experiments, with twin photons and coincidence detection, experiment using a nonlocal double slit. We have presented a
beams were also passed through localized interferometers, quantum theory in good agreement with experimental re-
such as a Mach-Zehnder 关16兴 or a Michelson 关18,19兴. sults. We have also pointed out the connections between our
In those experiments where localized interferometers experiment and other ones that violate Bell inequalities.
were used, it is possible to provide local interpretations. For
example, when intensity interference fringes do not show up
and coincidence interference fringes do, we can think of ACKNOWLEDGMENTS
them as being a result of a spectral filtering process due to
the coincidence measurements. This is a consequence of the The authors acknowledge financial support from the Bra-
fact that when SPDC light passes through a local interferom- zilian agencies CNPq, FINEP, PRONEX, and FAPEMIG.
PRA 60 QUANTUM INTERFERENCE BY A NONLOCAL DOUBLE SLIT 1533

关1兴 C. H. Bennett, Phys. Today 48 共10兲, 24 共1995兲. 关12兴 P. H. S. Ribeiro, S. Pádua, J. C. M. da Silva, and G. A. Bar-
关2兴 D. Bouwmeester et al., Nature 共London兲 390, 575 共1997兲. bosa, Phys. Rev. A 49, 4176 共1994兲.
关3兴 D. Boschi et al., Phys. Rev. Lett. 80, 1121 共1998兲. 关13兴 D. V. Strekalov, A. V. Sergienko, D. N. Klyshko, and Y. H.
关4兴 The expression ‘‘nonlocal’’ is too general in this context. Rig- Shih, Phys. Rev. Lett. 74, 3600 共1996兲.
orously speaking, we should use the more appropriate expres- 关14兴 C. H. Monken, P. H. S. Ribeiro, and S. Pádua, Phys. Rev. A
sion ‘‘nonlocal-realistic.’’ However, for the sake of brevity, 57, 3123 共1998兲.
many workers refer to ‘‘nonlocal-realistic’’ as ‘‘nonlocal.’’ In 关15兴 D. Greenberger, M. A. Horne, and A. Zeilinger, Phys. Today
this paper we adopt this common practice. 46 共8兲, 22 共1993兲.
关5兴 A. Einstein, B. Podolsky, and N. Rosen, Phys. Rev. 47, 777 关16兴 J. D. Franson, Phys. Rev. Lett. 62, 2205 共1989兲.
共1935兲. 关17兴 Z. Y. Ou, X. Y. Zou, L. J. Wang, and L. Mandel, Phys. Rev.
关6兴 A. Aspect, P. Grangier, and G. Roger, Phys. Rev. Lett. 47, 460
Lett. 65, 321 共1990兲.
共1981兲.
关18兴 P. G. Kwiat et al., Phys. Rev. A 41, 2910 共1990兲.
关7兴 Z. Y. Ou and L. Mandel, Phys. Rev. Lett. 61, 50 共1988兲.
关19兴 P. G. Kwiat and R. Y. Chiao, Phys. Rev. Lett. 66, 588 共1991兲.
关8兴 Y. H. Shih and C. O. Alley, Phys. Rev. Lett. 61, 2921 共1988兲.
关20兴 J. G. Rarity and P. R. Tapster, Phys. Rev. Lett. 64, 2495
关9兴 J. R. Torgerson, D. Branning, C. H. Monken, and L. Mandel,
共1990兲.
Phys. Lett. A 204, 323 共1995兲.
关10兴 J. R. Torgerson, D. Branning, C. H. Monken, and L. Mandel, 关21兴 L. Mandel and E. Wolf, Optical Coherence and Quantum Op-
Phys. Rev. A 51, 4400 共1995兲. tics 共Cambridge University Press, Cambridge, England, 1995兲.
关11兴 T. B. Pittman et al., Phys. Rev. A 53, 2804 共1996兲. 关22兴 Z. Y. Ou, Phys. Rev. A 37, 1607 共1988兲.

View publication stats