CoChat

Integrated photonic quantum walks

Open access
Gräfe, Markus, Szameit, Alexander · 2020 · Journal of Physics B: Atomic, Molecular and Optical Physics

The DOI matches its Crossref record; 166 of the 174 references it cites were found in a registry or an index; 24 of the 44 sources its claims cite were checked for support.

Abstract

Quantum walks (QWs) are one of the fundamental concepts in quantum information science, being an essential component for various new applications. Implementing QWs using integrated photonic waveguide lattices allows for realizing long evolution times, for exploring novel lattice geometries, and even for controllable coupling to the environment. This tutorial gives an introduction to the field of on-chip QWs of single photons and photon pairs featuring non-classical two-photon correlations.

Overview

CoChat Summary

TL;DR

This tutorial reviews integrated photonic quantum walks implemented in waveguide lattices, covering discrete- and continuous-time dynamics of single photons and photon pairs. It explains fabrication, photon-pair generation/detection, and nonclassical correlation phenomena including W-states, perfect state transfer, Anderson localization, Bloch oscillations, and non-Hermitian QWs.

Key findings

  • Integrated photonic lattices allow long evolution times, 3D geometries, and controllable coupling; direct laser-written waveguides have refractive-index increase ~10^-3, straight-section losses 0.3–0.9 dB cm^-1, and cm-scale bending radii, while silica-on-silicon offers 0.1 dB cm^-1 but only 2D structures.
  • In discrete-time single-photon QWs, a 31-step DC mesh shows quantum interference: after three steps the photon has vanishing probability in channel 3, and unlike the binomial Galton-board output, the walker most likely emerges opposite the input; adding random phases after each DC (uniform [0,π], averaged over 20,000 realizations) washes out ballistic lobes into a diffusive/binomial/Gaussian distribution.
  • Two-photon QWs exhibit nonclassical correlations: a 50:50 DC produces HOM interference with measured visibility approximately 96.4%; separable indistinguishable photons show bunching, path-entangled input shows anti-bunching, and distinguishable photons show four characteristic correlation peaks.
  • The tutorial reports on-chip W-state generation up to N=16, perfect state transfer in a 19-waveguide J_x lattice (input guide 1↔19, 2↔18), Anderson colocalization in 101-site disordered arrays with N=20 input, Bloch oscillations with revival success probability >92% in a 16-waveguide lattice, and non-Hermitian quasi-PT-symmetric lattices where transport becomes diffusive and row sums of the correlation function no longer equal average particle number.

Methods

  • Design: tutorial/review with theoretical derivations and discussion of experiments; introduces discrete-time QWs via directional couplers and continuous-time QWs via evanescently coupled waveguide arrays; uses Heisenberg and quantum Langevin equations, unitary evolution operator, and correlation functions.
  • Data/sample: no new primary dataset; draws on cited integrated-photonic experiments, including an AlGaAs 100-waveguide array, a 31-step DC mesh, a 5-waveguide W-state array, a 19-waveguide J_x lattice, 101-site disordered arrays, a 16-waveguide BO lattice, and non-Hermitian arrays. Photon pairs are generated by SPDC and detected by APDs, SNSPDs, or TESs.
  • Main analysis: analytical/numerical modeling of photon number and two-photon correlation functions, compared with experimental coincidence/correlation maps; nonclassicality is assessed via Bell-like inequalities and a visibility threshold (above 50% ensures nonclassical quantum interference).

Limitations

  • Stated in the paper: full on-chip integration of source, waveguide chip, and detection has not been realized; losses between source, chip, and detectors due to mode-size mismatch can be minimized with thermally enhanced core fibers and adiabatic inscription. Direct laser-written waveguides have moderate index increase (~10^-3), cm-scale bending radii, and 0.3–0.9 dB cm^-1 losses; silica-on-silicon is limited to 2D. Detection limitations include APDs cannot resolve multiple simultaneous photons and InGaAs telecom versions have enhanced noise/low efficiency; SNSPDs require <4 K; TESs require ~100 mK, have lower count rates, and photon-number discrimination degrades for higher N. The tutorial also states its generation/detection overview is brief and not essential to the principles.
  • Evident from the text: the manuscript is a tutorial/review rather than a primary research report, so its quantitative “findings” are syntheses of cited experiments rather than a single new experimental dataset.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1088/1361-6455/ab6cfc

Cites

166 of 166 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1 Sketch of a DC. Due to the 50:50 splitting ratio the integrated photonic waveguide structure transforms the input state ñ in ∣ into the two-mode path-entangled output state ñ out ∣ . Reproduced from [ 16 ] . © IOP Publishing Ltd. All rights reserved.
Figure 1 Sketch of a DC. Due to the 50:50 splitting ratio the integrated photonic waveguide structure transforms the input state ñ in ∣ into the two-mode path-entangled output state ñ out ∣ . Reproduced from [ 16 ] . © IOP Publishing Ltd. All rights reserved.CC-BY · From the PDF
Figure 2 Mesh of DCs building the fi rst three steps of a photonic discrete-time QW lattice. Reproduced from [ 16 ] . © IOP Publishing Ltd. All rights reserved.
Figure 2 Mesh of DCs building the fi rst three steps of a photonic discrete-time QW lattice. Reproduced from [ 16 ] . © IOP Publishing Ltd. All rights reserved.CC-BY · From the PDF
Figure 3 Discrete-time QW of single photons. A discrete lattice with 31 discrete evolution steps is formed by a mesh of DCs. ( a ) Singlephoton evolution when launched into one ( right ) input channel of a DC. ( b ) Single-photon evolution when prepared in a superposition state where both input ports of the fi rst DC are excited. ( c ) Same input state as in ( b ) but now with random phase shifts after each DC in the mesh. This dynamical disorder causes a diffusive evolution of the walker, which corresponds to the classical random walk scenario. The top row shows the corresponding photon density at the output. Reproduced from [ 16 ] . © IOP Publishing Ltd. All rights reserved.
Figure 3 Discrete-time QW of single photons. A discrete lattice with 31 discrete evolution steps is formed by a mesh of DCs. ( a ) Singlephoton evolution when launched into one ( right ) input channel of a DC. ( b ) Single-photon evolution when prepared in a superposition state where both input ports of the fi rst DC are excited. ( c ) Same input state as in ( b ) but now with random phase shifts after each DC in the mesh. This dynamical disorder causes a diffusive evolution of the walker, which corresponds to the classical random walk scenario. The top row shows the corresponding photon density at the output. Reproduced from [ 16 ] . © IOP Publishing Ltd. All rights reserved.CC-BY · From the PDF
Figure 4 Average photon number evolution nq ( z ) . A single photon is sent into the central channel of an array consisting of 40 identical evanescently coupled waveguides. The photon density evolves in the same way as the intensity when shining in coherent laser light with the same wavelength. Two typical ballistic lobes can be clearly seen.
Figure 4 Average photon number evolution nq ( z ) . A single photon is sent into the central channel of an array consisting of 40 identical evanescently coupled waveguides. The photon density evolves in the same way as the intensity when shining in coherent laser light with the same wavelength. Two typical ballistic lobes can be clearly seen.CC-BY · From the PDF
Figure 5 Integrated optical HOM effect. ( a ) Two photons enter a 50:50 DC and become spatially entangled. ( b ) Measurement of the twophoton coincidences at the DC output. By tuning the photons ' ( in ) distinguishability by their time delay the typical ' HOM-dip ' can be recorded. Here, the visibility ( drop in coincidences per average value ) is approximately 96.4%.
Figure 5 Integrated optical HOM effect. ( a ) Two photons enter a 50:50 DC and become spatially entangled. ( b ) Measurement of the twophoton coincidences at the DC output. By tuning the photons ' ( in ) distinguishability by their time delay the typical ' HOM-dip ' can be recorded. Here, the visibility ( drop in coincidences per average value ) is approximately 96.4%.CC-BY · From the PDF
Figure 6 Quantum correlations. Centered two-photon excitation at sites m = 20 and n = 21 in a homogeneously coupled array of 40 waveguides. ( a ) Average photon number evolution nq ( z ) . ( b ) Two-photon correlation matrix G q r , at the array output for the cases of both photons being indistinguishable and in a separable state, and ( c ) both photons entering in the path-entangled state + ñ a a 0 1 2 20 2 21 2 [( ˆ ) ( ˆ ) ]∣ † † . Clear photon bunching and anti-bunching patterns are shown, respectively. ( d ) The correlation map for the case of two distinguishable photons at the input exhibits the characteristic four peaks.
Figure 6 Quantum correlations. Centered two-photon excitation at sites m = 20 and n = 21 in a homogeneously coupled array of 40 waveguides. ( a ) Average photon number evolution nq ( z ) . ( b ) Two-photon correlation matrix G q r , at the array output for the cases of both photons being indistinguishable and in a separable state, and ( c ) both photons entering in the path-entangled state + ñ a a 0 1 2 20 2 21 2 [( ˆ ) ( ˆ ) ]∣ † † . Clear photon bunching and anti-bunching patterns are shown, respectively. ( d ) The correlation map for the case of two distinguishable photons at the input exhibits the characteristic four peaks.CC-BY · From the PDF
Figure 7 Direct laser-writing. ( a ) Microscope image of a laserwritten waveguide structure from the glass end facet. ( b ) Focusing fs-laser pulses into a glass sample permanently modi fi es the material in the focal volume resulting in a refractive index increase. By moving the sample through the incident beam, the waveguides are inscribed. Typical laser parameters are a center wavelength of 800 nm, a pulse width of 150 fs and a pulse energy of 500 μ J.
Figure 7 Direct laser-writing. ( a ) Microscope image of a laserwritten waveguide structure from the glass end facet. ( b ) Focusing fs-laser pulses into a glass sample permanently modi fi es the material in the focal volume resulting in a refractive index increase. By moving the sample through the incident beam, the waveguides are inscribed. Typical laser parameters are a center wavelength of 800 nm, a pulse width of 150 fs and a pulse energy of 500 μ J.CC-BY · From the PDF
Figure 8 Schematic image of a typical setup that is used for integrated photonic experiments.
Figure 8 Schematic image of a typical setup that is used for integrated photonic experiments.CC-BY · From the PDF
Figure 9 Coherent QW in a regular waveguide lattice. ( a ) Schematic of the one-dimensional ridge waveguide lattice, which was illuminated by a 1550 nm laser. ( b ) Evolution of the probability distribution ( simulation ) of a coherent QW. ( c ) Observed output ( intensity ) distribution when launching coherent laser light into the waveguide array as indicated in ( a ) . The results are obtained at short ( blue ) and long ( green ) propagation distances marked by the dashed lines in ( b ) . Reprinted fi gure with permission from [ 68 ] , Copyright 2008 by the American Physical Society.
Figure 9 Coherent QW in a regular waveguide lattice. ( a ) Schematic of the one-dimensional ridge waveguide lattice, which was illuminated by a 1550 nm laser. ( b ) Evolution of the probability distribution ( simulation ) of a coherent QW. ( c ) Observed output ( intensity ) distribution when launching coherent laser light into the waveguide array as indicated in ( a ) . The results are obtained at short ( blue ) and long ( green ) propagation distances marked by the dashed lines in ( b ) . Reprinted fi gure with permission from [ 68 ] , Copyright 2008 by the American Physical Society.CC-BY · From the PDF
Figure 10 Single-photon evolution in tailored waveguide arrays. Average photon number when exciting the central of fi ve waveguides and tailored coupling strengths κ 1 and κ 2. An equal density distribution is achieved at the propagation distance z W , which is marked by a dashed green line.
Figure 10 Single-photon evolution in tailored waveguide arrays. Average photon number when exciting the central of fi ve waveguides and tailored coupling strengths κ 1 and κ 2. An equal density distribution is achieved at the propagation distance z W , which is marked by a dashed green line.CC-BY · From the PDF
Figure 11 Generation of high-order single-photon W-states on-chip. ( a ) Five continuously coupled waveguides with tailored coupling strengths for generation of ñ W5 ∣ . ( b ) Scheme resting upon discrete splitting steps implemented via balanced DCs. ( c ) Measured probability statistics for detecting a photon at the output for several orders of W-states. The distributions are close to uniform and their standard deviations ( compared to ideal uniform distributions ) are shown as horizontal bars. The generation of ñ W16 ∣ is not discussed here, but can be seen as a straightforward extension of the scheme. Reprinted by permission from Macmillan Publishers Ltd: Nat Photon [ 97 ] , Copyright 2014.
Figure 11 Generation of high-order single-photon W-states on-chip. ( a ) Five continuously coupled waveguides with tailored coupling strengths for generation of ñ W5 ∣ . ( b ) Scheme resting upon discrete splitting steps implemented via balanced DCs. ( c ) Measured probability statistics for detecting a photon at the output for several orders of W-states. The distributions are close to uniform and their standard deviations ( compared to ideal uniform distributions ) are shown as horizontal bars. The generation of ñ W16 ∣ is not discussed here, but can be seen as a straightforward extension of the scheme. Reprinted by permission from Macmillan Publishers Ltd: Nat Photon [ 97 ] , Copyright 2014.CC-BY · From the PDF
Image 12
CC-BY · From the PDF
Figure 13 Anderson co-localization -experimental setup. Spatially entangled photon pairs are generated by SPDC in a nonlinear crystal ( NLC ) . Lens L1 couples them into the waveguide array. Spatial correlations in the output fi eld are measured using two scanning fi ber tips ( planes x 1 and x 2 ) , which are connected to single-photon detectors ( SPCM1 and SPCM2 ) . This is realized by another lens L2 and a beam splitter BS. The actual photonic chip used in the experiment is shown in the inset. Reprinted fi gure with permission from [ 117 ] , Copyright 2013 by the American Physical Society.
Figure 13 Anderson co-localization -experimental setup. Spatially entangled photon pairs are generated by SPDC in a nonlinear crystal ( NLC ) . Lens L1 couples them into the waveguide array. Spatial correlations in the output fi eld are measured using two scanning fi ber tips ( planes x 1 and x 2 ) , which are connected to single-photon detectors ( SPCM1 and SPCM2 ) . This is realized by another lens L2 and a beam splitter BS. The actual photonic chip used in the experiment is shown in the inset. Reprinted fi gure with permission from [ 117 ] , Copyright 2013 by the American Physical Society.CC-BY · From the PDF
Figure 14 Anderson co-localization -results. Columns: correlation functions for a spatially entangled two-photon state in periodic and disordered waveguide arrays along ( a ) periodic, ( b ) weakly disordered, and ( c ) highly disordered waveguide lattices. The propagation distance is L = 5 cm. Rows: ( i ) calculated as well as ( ii ) measured coincidence rates G x x , 2 1 2 ( ) ( ) , ( iii ) single-photon events G x x , 1 1 1 ( ) ( ) , and ( iv ) distribution of the diagonals. Reprinted fi gure with permission from [ 117 ] , Copyright 2013 by the American Physical Society.
Figure 14 Anderson co-localization -results. Columns: correlation functions for a spatially entangled two-photon state in periodic and disordered waveguide arrays along ( a ) periodic, ( b ) weakly disordered, and ( c ) highly disordered waveguide lattices. The propagation distance is L = 5 cm. Rows: ( i ) calculated as well as ( ii ) measured coincidence rates G x x , 2 1 2 ( ) ( ) , ( iii ) single-photon events G x x , 1 1 1 ( ) ( ) , and ( iv ) distribution of the diagonals. Reprinted fi gure with permission from [ 117 ] , Copyright 2013 by the American Physical Society.CC-BY · From the PDF
Image 15
CC-BY · From the PDF
Figure 15 Anderson localization of quantum states when propagating in a lattice with on-diagonal disorder. ( a ) Discrete network of DCs. ( b ) Two-photon correlations for ( from top to bottom ) eight, 12, and 16 steps. The input state is y ñ 0 ∣ ( ) . ( c ) The same as ( b ) for the input state y ñ p ∣ ( ) . ( d ) Average particle number obtained by tracing out the position of one particle ( column by column summation of the correlation map ) . Reprinted by permission from Springer Nature Customer Service Centre GmbH: Springer Nature, Nature Photonics [ 118 ] 2013.
Figure 15 Anderson localization of quantum states when propagating in a lattice with on-diagonal disorder. ( a ) Discrete network of DCs. ( b ) Two-photon correlations for ( from top to bottom ) eight, 12, and 16 steps. The input state is y ñ 0 ∣ ( ) . ( c ) The same as ( b ) for the input state y ñ p ∣ ( ) . ( d ) Average particle number obtained by tracing out the position of one particle ( column by column summation of the correlation map ) . Reprinted by permission from Springer Nature Customer Service Centre GmbH: Springer Nature, Nature Photonics [ 118 ] 2013.CC-BY · From the PDF
Figure 16 Integrated optical waveguide structure for experimental implementation of two-photon N00N state BOs. The curved center part is the Bloch oscillator lattice. It consists of 16 evanescently coupled waveguides and is directly linked to a detuned DC, which performs the state preparation. A fan-out section at the end is for technical reasons, to connect to fi ber arrays guiding the photons to single-photon detectors. Reproduced from [ 121 ] . CC BY 4.0.
Figure 16 Integrated optical waveguide structure for experimental implementation of two-photon N00N state BOs. The curved center part is the Bloch oscillator lattice. It consists of 16 evanescently coupled waveguides and is directly linked to a detuned DC, which performs the state preparation. A fan-out section at the end is for technical reasons, to connect to fi ber arrays guiding the photons to single-photon detectors. Reproduced from [ 121 ] . CC BY 4.0.CC-BY · From the PDF
Figure 17 BOs of symmetric path-entangled states. Experimentally obtained and theoretically calculated correlation maps and the corresponding inter-particle distances ( top to bottom ) for evolution lengths l ¼ 0.1 0.5 B ( ) ( left to right ) . Reproduced from [ 121 ] . CC BY 4.0.
Figure 17 BOs of symmetric path-entangled states. Experimentally obtained and theoretically calculated correlation maps and the corresponding inter-particle distances ( top to bottom ) for evolution lengths l ¼ 0.1 0.5 B ( ) ( left to right ) . Reproduced from [ 121 ] . CC BY 4.0.CC-BY · From the PDF
Figure 18 Non-Hermitian QW in quasi- -symmetric lattices. ( a ) Analyzed waveguide array with alternating lossy channels ( red ) and loss-free guides ( blue ) with the corresponding loss pro fi le below. Two-photon correlation matrices for various input con fi gurations: separable input state launched in ( b ) adjacent, ( c ) next-nearest neighboring lossy, and ( d ) next-nearest neighboring loss-free sites. Coupling between adjacent sites was 1 cm -1 and loss 4 cm -1 . Reproduced from [ 24 ] . © IOP Publishing Ltd. CC BY 3.0.
Figure 18 Non-Hermitian QW in quasi- -symmetric lattices. ( a ) Analyzed waveguide array with alternating lossy channels ( red ) and loss-free guides ( blue ) with the corresponding loss pro fi le below. Two-photon correlation matrices for various input con fi gurations: separable input state launched in ( b ) adjacent, ( c ) next-nearest neighboring lossy, and ( d ) next-nearest neighboring loss-free sites. Coupling between adjacent sites was 1 cm -1 and loss 4 cm -1 . Reproduced from [ 24 ] . © IOP Publishing Ltd. CC BY 3.0.CC-BY · From the PDF

Verification

Claims the paper makes

13 Valid · 4 Not valid · 15 Not checked · 44 Own results and interpretations

33 more claims · 1 with their cited sources checked

Create a free account to see all 76 claims and every source verdict.

See all claims — free account
  • Probably not a substantial claim. 19
  • It may state two separate claims as one. 25
  • The quote only partly states this claim. 2

Create a free account to see these claims.

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

3 of 3

Citations

Cites

Research works printed but not linked: 1

  1. Yariv A 1989 Quantum Electronics (New York: Wiley)

Cited by

See all 28

Metadata

Authors
Gräfe, M., & Szameit, A.
Affiliations
  • Fraunhofer Institute for Applied Optics and Precision Engineering (DE)
  • University of Rostock (DE)
Corresponding author
Gräfe, Markus
Venue
Journal of Physics B: Atomic, Molecular and Optical Physics
Published
2020-03-10
Volume
53
Issue
7
Pages
073001
ISSN
  • 0953-4075 (ISSN-L)
  • 1361-6455
OpenAlex
W2999198815
Cited by
28
References
174
Citations by year
2025: 5 · 2024: 5 · 2023: 5 · 2022: 4 · 2021: 9
FWCI
1.74
Type
journal article
Language
en
Open access
Licence
Locations
  • Journal of Physics B Atomic Molecular and Optical Physics (publishedVersion, cc-by)
  • Publikationsdatenbank der Fraunhofer-Gesellschaft (Fraunhofer-Gesellschaft) (submittedVersion)
  • Fraunhofer-Publica (Fraunhofer-Gesellschaft) (submittedVersion)
Funders
  • Deutsche Forschungsgemeinschaft (SZ 276/12-11)
  • Fraunhofer-Gesellschaft (QUILT)
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Quantum Computing Algorithms and Architecture
  • Physical Sciences › Physics and Astronomy › Atomic and Molecular Physics, and Optics › Quantum Information and Cryptography
  • Physical Sciences › Physics and Astronomy › Atomic and Molecular Physics, and Optics › Quantum optics and atomic interactions
Keywords
quantum walks, integrated photonics, single photons, photon pairs, two-photon correlations, nonclassical correlations

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-07, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-07 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.