Franson-type nonlocal correlation results in a second-order intensity fringe between two remotely separated parties via coincidence measurements, whereas the corresponding local measurements show a perfect incoherence feature. This nonlocal correlation fringe between paired photons is mysterious due to the local randomness in both parties. Here, the Franson nonlocal correlation fringe is analytically investigated using the wave nature of photons to understand the mysterious quantum feature. As a result, the nonlocal intensity fringe is turned out to be a measurement selection-based coherence feature, while the local randomness is from effective decoherence among broad bandwidth-distributed photon pairs. As a result, a coherence version of Franson nonlocal correlation is suggested for macroscopic quantum applications with a commercial laser. The local and nonlocal correlations of the proposed scheme show the same results as entangled photon-pair based Franson correlation.

At the request of the authors, this article is being retracted effective 28 October 2022.

Quantum entanglement is known as a weird phenomenon that cannot be explained by classical physics or achieved by any classical means.1 Ever since the well-known thought experiment by Einstein, Podolsky, and Rosen (EPR) in 1935,2 EPR has been a key aspect of quantum information science and technologies in computing,3–5 communications,6–9 and sensing areas.10–12 Although Bell had mathematically demonstrated the so-called EPR paradox in 1964, the definition of classical physics rules out coherence optics.13 Since then, mutual coherence between paired photons has not been carefully considered.13–26 Although the energy-time uncertainty relation prohibits definite phase information for a single photon, a mutual phase between entangled photons is free from such uncertainty relation without violating quantum mechanics. Here, we focus on the mutual coherence between interacting photons in Franson-type nonlocal correlation to understand the mysterious quantum feature. The classical understanding stands for a definite and clear reasoning-based physical nature.

Franson-type nonlocal correlation19 has been studied since 1987.20–26 Unlike Bell inequality violations,13–18 Franson nonlocal correlation is based on a set of unbalanced Mach–Zehnder interferometers (U-MZIs).19–26 The U-MZI is designed with respect to interacting photons' bandwidth for both local randomness in each detector and nonlocal correlation fringe in coincidence detection. The local randomness-based nonlocal fringe is a unique feature of the Franson correlation. Although the mathematical form of the nonlocal correlation is clear to view the fringe, the physical understanding of the nonlocal fringe has been severely limited. Thus, the fringe in Franson nonlocal correlation has been left as a mysterious quantum feature due to the U-MZI resulting local randomness. Here, the origin of Franson nonlocal correlation is investigated to clear out the weirdness of nonlocal fringe based on local randomness. For this, we analyze the relation between U-MZI and entangled photon pairs generated from the spontaneous parametric down conversion (SPDC) process.27–31 Specifically, the coherence relation between each entangled photon pair and the U-MZI is focused on for coincidence detection. From this understanding, a coherence version of the Franson correlation is proposed for macroscopic quantum information compatible with current technologies based on coherence optics.

Figure 1 shows the original Franson scheme for nonlocal correlation using entangled photon pairs generated from SPDC processes.20 As studied29 and applied for quantum key distributions,24–26 the coincidence measurements between two remotely separated output photons show a path-length difference-dependent fringe, even though their local measurements do not. This nonlocal fringe for the second-order intensity correlation looks exactly the same as the first-order intensity correlation in a typical double-slit case. Regarding the U-MZIs in Fig. 1(a), whose entangled photon source S is depicted in Figs. 1(b) and 1(c), the nonlocal correlation fringe due to the coincidence measurements implies that each U-MZI may act as a coherence system for individual photon pairs. Considering this, the coherence time τ c of each entangled photon pair from the SPDC should be much longer than the path-length difference δ L L S of each U-MZI. This coherence relation can be achieved by choosing a narrow bandwidth pump laser (p) in Fig. 1(c) according to χ ( 2 ) nonlinear optics of SPDC.20,30,31 As explained below in Eq. (11), the coherence washout among the broadband photon pairs is suppressed by coincidence measurements. Due to the ultrawide bandwidth ( Δ ) of the entangled photon ensemble in Fig. 1(b), however, each U-MZI in Fig. 1(a) acts as a noninterfering interferometer for local measurements, resulting in the path-length independent uniform intensity. Thus, both local and nonlocal measurements are easily understood by many-wave interference in coherence optics.

Fig. 1.

Schematic of Franson-type nonlocal correlation. (a) The original Franson setup. (b) Probability distribution of the frequency of the light source S in (a). (c) Schematic of type I SPDC. 2 f 0 = f s + f i. S (short path) and L (long path) indicate the two paths of each unbalanced MZI. D: single photon detector. In (b), the full width at half maximum of entangled photons is Δ. In (c), p, s, and i indicate pump, signal, and idler photons, respectively, where s and i can be swapped due to the phase matching condition (see the text).

Fig. 1.

Schematic of Franson-type nonlocal correlation. (a) The original Franson setup. (b) Probability distribution of the frequency of the light source S in (a). (c) Schematic of type I SPDC. 2 f 0 = f s + f i. S (short path) and L (long path) indicate the two paths of each unbalanced MZI. D: single photon detector. In (b), the full width at half maximum of entangled photons is Δ. In (c), p, s, and i indicate pump, signal, and idler photons, respectively, where s and i can be swapped due to the phase matching condition (see the text).

Close modal
For the U-MZI (Alice) in Fig. 1(a), coherence optics for each input photon E 0 results in the following matrix representation for the output photons:
(1)
where B S = 1 2 1 i i 1, φ j = 1 0 0 e i φ j , and φ j = δ f j τ + φ. Here, the amplitude of a single photon is set at E 0, where a harmonic oscillation term is omitted for simplicity. τ is the temporal difference of the photon between L and S of the U-MZI, τ = L S / c, where c is the speed of light. δ f j is the detuning of the jth photon pair with respect to the center frequency f 0 as shown in Fig. 1(b).

Figure 1(c) shows schematic of the type I SPDC for the same polarization of entangled photon pairs. Due to the phase matching condition of SPDC governed by both energy and momentum conservation laws,30,31 the signal (s) and idler (i) photons are interchangeable, satisfying an entangled state, | ψ = 1 2 | s 1 | i 2 + | i 1 | s 2, where the subscripts indicate concentric circles. Thus, the f 0 in Fig. 1(b) is the half of the pump photon frequency, whose spectral width is ultranarrow compared with the generated photon bandwidth Δ. Here, having a narrow linewidth pump laser is essential for the higher efficiency of nonlocal fringe visibility via stable concentric ring patterns in Fig. 1(c) according to the momentum conservation law.20,31,32 According to the phase matching in SPDC nonlinear optics,31 the paired (signal and idler) photons satisfy symmetric frequency detuning ( δ f j) across the center frequency f 0, resulting in a ± δ f j relation with 2 f 0 = f s + f i according to the energy conservation law.31 

Likewise, the matrix representation for Bob's side in Fig. 1(a) is as follows:
(2)
where ψ j = δ f j τ + ψ. As a result, the mean amplitudes of each output port in both sides are as follows:
(3)
(4)
(5)
(6)
In Eqs. (3)–(6), the superscript A (B) indicates Alice (Bob), and the subscript 5 (6) is for the output port number. From Eqs. (3)–(6), the corresponding mean intensities are as follows:
(7)
(8)
(9)
(10)
Thus, the ensemble average of local measurements becomes uniform at I 0 / 2 due to wide bandwidth Δ, resulting the local randomness. This uniform intensity in local detections is, of course, due to many-wave interference, satisfying the incoherence optics of U-MZIs. As a result, this no-fringe in local measurements can be easily understood as a direct result of coherence optics for photon ensemble average.
The mean coincidence measurements between remotely separated output photons, e.g., R 5 j A B ( = I 5 j A I 5 j B ), are directly calculated from the product of equations (7) and (9):
(11)
In Eq. (11), the last two terms are related to the product of the coincidence detection. By definition of coincidence ( τ = 0 ), the last two terms in the bracket become equal for ψ = 0, regardless of δ f j. Here, the effective coherence time τ c is Δ 1. The τ c-based coherence length l c is much longer than the wavelength λ of each entangled photon. For coincidence detection ( τ = 0 ), Eq. (11) turns out to be φ-dependent (for ψ = 0),
(12)
Equation (12) is the heart of Franson-type nonlocal correlation, where the fringe depends only on the path-length variation δ L ( = L B L A ) in U-MZIs. Due to the fine phase controllability of U-MZIs with δ L, the bracket of the cos φ term is removed. This path-length dependent fringe in nonlocal correlation is originated in the particular measurement process of coincidence detection ( τ = 0 ), otherwise results in the classical lower bound, R 55 A B = I 0 2 / 4 1 + 1 / 2  cos  φ. This is the coherence interpretation of the Franson-type nonlocal correlation in the present analysis.
Likewise, all other nonlocal correlations between the paired photons coincidently measured for ψ = 0 in both parties are as follows:
(13)
(14)
(15)
Equations (12)–(15) are exactly the same as those observed in Ref. 32 for loophole-free Franson correlation. Thus, the phase φ in Eqs. (11)–(15) is nothing but for the path-length difference between two parties, δ L ( = L B L A ). As a result, Franson-type nonlocal correlation can be considered as a special version of coherence optics via coincidence measurements. General coherence optics without a temporal (spectral) modification of the coincidence detection cannot result in this nonlocal property.

Over the last several decades, Franson-type nonlocal correlation has been applied to quantum key distributions based on the energy-time bin method using single photon-correlated entangled pairs.24–26 Because this nonlocal phenomenon is successfully explained by coherence optics for coincidence measurements in Eq. (12), a coherence version of the Franson-type nonlocal correlation can be considered as well. In quantum mechanics, the believed thumb rule is that nonlocal correlation cannot be achieved by any classical means. Thus, the present analysis provides some insight into the quantum nature.

Figure 2 shows a coherence version of Fig. 1, where the entangled photon pairs are replaced by coherent photons from a commercial laser with some modifications. Unlike the SPDC-generated entangled photon pairs, the spectral bandwidth Δ l of each coherent photon is the same as the laser linewidth according to cavity optics.33 The function of an optical cavity is for spectral filtering of wide-bandwidth distributed photons generated from a gain medium. To satisfy both temporal and spectral integrations for nonlocal and local measurements, respectively, the laser-generated photons are modified to be inhomogeneous for each U-MZI in Fig. 2. This inhomogeneity of coherent photons results in the same function of overall decoherence in ensemble measurements of SPDC. Such spectral widening of coherent photons can be accomplished by laser scanning, such as in frequency modulation of continuous waves34 or dc Stark effects.35 

Fig. 2.

Schematic of a coherence version of Franson nonlocal correlation. AOM: acousto-optic modulator; D: photodetector; SM: spectral modifier.

Fig. 2.

Schematic of a coherence version of Franson nonlocal correlation. AOM: acousto-optic modulator; D: photodetector; SM: spectral modifier.

Close modal
The symmetric frequency relation between the signal and idler photons in Fig. 1(b) is satisfied through the use of acousto-optic modulators (AOMs) in Fig. 2. For this, both AOMs are driven by a common microwave field ( f r f ) to generate a frequency-shifted light pair, whose phase is synchronized. The symmetric frequency configuration between the paired fields is accomplished by choosing oppositely diffracted lights. Figure 2 shows the negative (positive) diffraction choice for Alice (Bob), resulting in f Alice = f 0 f r f + Δ l ( f Bob = f 0 + f r f + Δ l ). The measurement product between two local detectors corresponding to Eq. (11) is as follows:
(16)
where φ j = δ f τ + φ, ψ j = δ f τ + ψ, and δ f = f r f + Δ l. By post-selection of the measurements corresponding to coincidence detection in Fig. 1, Eq. (16) becomes
(17)
Equation (17) is exactly the same as Eq. (12) for ψ = 0, regardless of the laser linewidth, satisfying the same nonlocal correlation as in a coherence regime. In this case, all is classical violating the common belief in nonlocal quantum features. In a short conclusion, the nonlocal property of Franson-type correlation has been viewed as a coincidence detection-caused modification (measurement filtering) of coherence optics. This thought experiment considering the coherence version of nonlocality relies on pure coherence optics in terms of the wave nature of photons in quantum mechanics. In that sense, quantum correlation does not have to be weird or mysterious but can be achieved by coherence optics, which is compatible with current optic technologies.

Franson-type nonlocal correlation was analyzed and discussed using the wave nature of photons for both local randomness and nonlocal correlation fringe. For this, the original unbalanced MZI (U-MZI) is investigated with respect to the characteristics of entangled photons generated from type I SPDC. From this, the locally measured uniform intensity in each detector was analyzed as a dephasing effect due to many-wave interference of spectrally broadened photons. On the contrary, the U-MZI path-length dependent intensity fringe for the nonlocal correlation was analyzed as a coherence feature of each photon pair in the U-MZI, where the U-MZI is designed to be coherent with respect to each photon pair. This coherence condition is satisfied with a narrow-bandwidth pump laser via coincidence detection. Unlike local measurements in time averaging, resulting in overall decoherence, the nonlocal measurements require coincidence detection between two remotely separated local measurements. This coincidence detection is key to understand the nonlocal fringe due to ruling out the time average effect for all spectral bandwidth entangled photons. Thus, the origin of both local randomness and nonlocal Franson correlation was found in coherence optics, where the nonlocal property is due to measurement filtering via coincidence detections. Based on this understanding, a coherence version of Franson nonlocal correlation was proposed using spectral modification of a commercially available laser light for the local randomness. The nonlocal correlation was achieved using a pair of acousto-optic modulators mimicking the SPDC-generated signal and idler photon pairs in symmetric frequency detuning.

This work was supported by the ICT R&D program of MSIT/IITP (No. 2021-0-01810) via Development of Elemental Technologies for ultra-secure Quantum Internet.

The authors have no conflicts to disclose.

B.S.H. solely wrote the manuscript.

Data sharing is not applicable to this article as no new data were created or analyzed in this study.

1.
N.
Brunner
et al,
Rev. Mod. Phys.
86
,
419
(
2014
).
2.
A.
Einstein
,
B.
Podolsky
, and
N.
Rosen
,
Phys. Rev.
47
,
777
(
1935
).
4.
H.-S.
Zhong
et al,
Phys. Rev. Lett.
127
,
180502
(
2021
).
7.
B.
Korzh
,
C. C. W.
Lim
,
R.
Houlmann
,
N.
Gisin
,
M. J.
Li
,
D.
Nolan
,
B.
Sanguinetti
,
R.
Thew
, and
H.
Zbinden
,
Nat. Photonics
9
,
163
(
2015
).
8.
P.
Sibson
et al,
Nat. Commun.
8
,
13984
(
2017
).
9.
I.
Marcikic
,
H.
de Riedmatten
,
W.
Tittel
,
H.
Zbinden
,
M.
Legre
, and
N.
Gisin
,
Phys. Rev. Lett.
93
,
180502
(
2004
).
10.
V.
Giovannetti
,
S.
Lloyd
, and
L.
Maccone
,
Nat. Photonics
5
,
222
(
2011
).
11.
C. L.
Degen
,
F.
Reinhard
, and
P.
Cappellaro
,
Rev. Mod. Phys.
89
,
035002
(
2017
).
12.
J. P.
Dowling
,
J. Lightwave Technol.
33
,
2359
2370
(
2015
).
14.
J. F.
Clauser
,
M. A.
Horne
,
A.
Shimony
, and
R. A.
Holt
,
Phys. Rev. Lett.
23
,
880
(
1969
).
15.
X.-S.
Ma
,
A.
Qarry
,
J.
Kofler
,
T.
Jennewein
, and
A.
Zeilinger
,
Phys. Rev. A
79
,
042101
(
2009
).
17.
A.
Aspect
,
P.
Grangier
, and
G.
Roger
,
Phys. Rev. Lett.
49
,
91
(
1982
).
18.
M.
Żukowski
,
A.
Zeilinger
,
M. A.
Horne
, and
A. K.
Ekert
,
Phys. Rev. Lett.
71
,
4287
(
1993
).
19.
20.
P. G.
Kwiat
,
A. M.
Steinberg
, and
R. Y.
Chiao
,
Phys. Rev. A
47
,
R2472
(
1993
).
21.
S.
Aerts
,
P.
Kwiat
,
J.-Å.
Larsson
, and
M.
Żukowski
,
Phys. Rev. Lett.
83
,
2872
(
1999
).
22.
A.
Cabello
,
A.
Rossi
,
G.
Vallone
,
F.
De Martini
, and
P.
Mataloni
,
Phys. Rev. Lett.
102
,
040401
(
2009
).
23.
G.
Carvacho
et al,
Phys. Rev. Lett.
115
,
030503
(
2015
).
24.
W.
Tittel
,
J.
Brendel
,
H.
Zbinden
, and
N.
Gisin
,
Phys. Rev. Lett.
81
,
3563
(
1998
).
25.
A.
Boaron
et al,
Appl. Phys. Lett.
112
,
171108
(
2018
).
26.
A.
Cuevas
et al,
Nat. Commun.
4
,
2871
(
2013
).
27.
R. W.
Boyd
,
Nonlinear Optics
, 3rd ed. (
Academic
,
New York
,
2008
), pp.
79
88
.
28.
H.
Defienne
and
S.
Gigan
,
Phys. Rev. A
99
,
053831
(
2019
).
29.
O.
Kwon
,
Y.-S.
Ra
, and
Y.-H.
Kim
,
Opt. Express
17
,
13059
(
2009
).
30.
B. E. A.
Saleh
and
M. C.
Teich
,
Fundamentals of Photonics
, 2nd ed. (
Wiley
,
New York
,
2012
), Chap. 21.
31.
H.
Cruz-Ramirez
,
R.
Ramirez-Alarcon
,
M.
Corona
,
K.
Garay-Palmett
, and
A. B.
U'Ren
,
Opt. Photonics News
22
,
36
41
(
2011
), and reference therein.
32.
G.
Lima
,
, G.
Vallone
,
A.
Chiuri
,
A.
Cabello
, and
P.
Mataloni
,
Phys. Rev. A
81
,
040101(R)
(
2010
).
33.
S.
Kim
and
B. S.
Ham
, arXiv:2104.08007 (
2021
).
34.
B.
Behroozpour
,
P.
Sandborn
,
M.
Wu
, and
B. E.
Boser
,
IEEE Commun. Mag.
55
,
135
142
(
2017
).
35.
A. L.
Alexander
,
J. J.
Longdell
,
M. J.
Sellars
, and
N. B.
Manson
,
Phys. Rev. Lett.
96
,
043602
(
2006
).