Kinetic Event-Chain Algorithm for Active Matter

Jan Kierfeld; Nico Schaffrath; Thevashangar Sathiyanesan; Tobias A. Kampmann

arxiv: 2111.06760 · v2 · submitted 2021-11-12 · ❄️ cond-mat.soft · physics.comp-ph

Kinetic Event-Chain Algorithm for Active Matter

Nico Schaffrath , Thevashangar Sathiyanesan , Tobias A. Kampmann , Jan Kierfeld This is my paper

Pith reviewed 2026-05-24 13:26 UTC · model grok-4.3

classification ❄️ cond-mat.soft physics.comp-ph

keywords active matterevent-chain Monte Carlokinetic Monte Carloactive Brownian particlesmotility-induced phase separationphase diagramsteric interactions

0 comments

The pith

A kinetic event-chain Monte Carlo algorithm simulates active matter correctly while running much faster than Brownian dynamics.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper presents a kinetic version of the event-chain Monte Carlo method for active particles with steric interactions. It determines a mean time for each move by comparing to Brownian dynamics, enabling correct rotation of the self-propulsion direction. This reproduces dynamics from single particles to collective phase separation in two-dimensional active disks. The method allows simulations of very large systems on ordinary computers.

Core claim

The kinetic event-chain algorithm correctly and efficiently reproduces physical results ranging from single-particle dynamics to many-body effects. In particular, it reproduces the phase diagram of active disks and the motility-induced phase separated region with high accuracy while being much faster than event-driven Brownian dynamics algorithms at comparable accuracy.

What carries the argument

The kinetic event-chain Monte Carlo method, which assigns a mean time to Monte Carlo moves derived from Brownian dynamics comparison to enable diffusional rotation of the propulsion force.

Load-bearing premise

The mean time assigned to Monte-Carlo moves accurately captures the diffusional rotation of the propulsion force and preserves correct active particle trajectories and collective statistics.

What would settle it

Running the algorithm on a standard model of active disks and finding that the reproduced phase diagram or single-particle trajectories deviate substantially from established Brownian dynamics results.

Figures

Figures reproduced from arXiv: 2111.06760 by Jan Kierfeld, Nico Schaffrath, Thevashangar Sathiyanesan, Tobias A. Kampmann.

**Figure 2.** Figure 2: FIG. 2. Simulation results in the eigenframe of the passive particle moving in positive x-direction. (a) Snapshot at [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** Figure 3: FIG. 3. Mean ECMC move lengths in force direction [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗

**Figure 4.** Figure 4: FIG. 4. (a) Double logarithmic plot of dimensionless sedimen [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗

**Figure 6.** Figure 6: FIG. 6. Results of additional simulations with phase sep [PITH_FULL_IMAGE:figures/full_fig_p009_6.png] view at source ↗

**Figure 7.** Figure 7: FIG. 7. (a) Snapshot of a system exhibiting MIPS at [PITH_FULL_IMAGE:figures/full_fig_p010_7.png] view at source ↗

**Figure 8.** Figure 8: FIG. 8. (a): Rescaled mean-squared displacements per simulation step as a function of simulations steps [PITH_FULL_IMAGE:figures/full_fig_p011_8.png] view at source ↗

**Figure 9.** Figure 9: FIG. 9. Normalized mean local packing fractions [PITH_FULL_IMAGE:figures/full_fig_p012_9.png] view at source ↗

read the original abstract

We present a cluster kinetic Monte-Carlo algorithm for active matter systems of self-propelled particles with special focus on steric interactions. The kinetic event-chain algorithm is based on the event-chain Monte-Carlo method and is applied to active Brownian disks in two dimensions. The algorithm assigns Monte-Carlo moves of active disks a mean time based on a comparison between Brownian dynamics and the dynamics of the event-chain Monte-Carlo method. This time is used to perform diffusional rotation of their propulsion force. We show that the algorithm correctly and efficiently reproduces various physical results ranging from single-particle dynamics to many-body-effects. In particular, we reproduce the phase diagram of active disks and the motility-induced phase separated region with high accuracy. The kinetic event-chain algorithm is shown to be much faster - at comparable accuracy - than (event-driven) Brownian dynamics algorithms, enabling large-scale simulations up to giant systems with $10^5$ particles on standard desktop hardware.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The kinetic event-chain adaptation for active disks offers a plausible route to larger simulations, but the mean-time calibration for rotational diffusion needs explicit checks in dense phases to confirm it holds.

read the letter

The new piece here is the specific way they map event-chain Monte Carlo moves for active particles onto a mean time that then drives the diffusional rotation of the propulsion vector. They derive that time by matching the event-chain dynamics to Brownian dynamics, then use it to evolve the orientation. This lets them keep the efficiency of event chains while adding the active rotation that standard ECMC lacks. If it works, the payoff is clear: they report running 10^5-particle systems on ordinary hardware where event-driven BD would be slower at the same accuracy. They also claim the method recovers both single-particle motion and the full phase diagram, including the MIPS region, without obvious artifacts. That combination of scale and claimed fidelity is the practical contribution. The soft spot is exactly the one the stress test flags. The mean time comes from a comparison whose regime is not stated to be universal. In the dense MIPS phase, particles spend more time in contact, so steric torques or caging could shift the effective rotational diffusion away from the dilute calibration value. The abstract asserts high accuracy but gives no error bars, no density-dependent tests, and no table showing how the assigned time was validated across packing fractions. Without those, it is possible the agreement is good only where the calibration was performed and drifts elsewhere. The paper does not appear to introduce new fitted parameters beyond that single mean time, which is a plus, but the robustness of that choice is the load-bearing assumption. This work is aimed at people who already run large active-matter simulations and want a faster alternative to BD. A reader who needs to simulate 10^4–10^5 particles to see collective effects would get the most out of it, provided the calibration holds. It is worth sending to referees because the algorithmic idea is concrete and the claimed performance gain is large enough to matter if the validation is solid; a referee can ask for the missing cross-checks on rotational diffusion in the dense phase.

Referee Report

2 major / 1 minor

Summary. The manuscript introduces a kinetic event-chain Monte Carlo algorithm for active Brownian disks in 2D with steric interactions. Monte Carlo moves are assigned a mean time derived from direct comparison to Brownian dynamics; this time drives diffusional rotation of each particle's propulsion vector. The central claim is that the resulting dynamics correctly reproduce single-particle trajectories, many-body effects, the full phase diagram of active disks, and the motility-induced phase separation (MIPS) region with high accuracy, while being substantially faster than event-driven Brownian dynamics and enabling simulations of 10^5 particles.

Significance. If the calibration procedure is shown to be robust, the method would provide a computationally efficient route to large-scale active-matter simulations that are currently limited by the cost of Brownian dynamics. The extension of event-chain Monte Carlo to include rotational diffusion via a calibrated mean time is a technically interesting contribution that could be useful for other active systems.

major comments (2)

[Abstract] Abstract: the claim that the algorithm reproduces the phase diagram and MIPS region 'with high accuracy' is unsupported by any quantitative error metrics, binodal comparisons, or validation details; without these, the central claim cannot be evaluated.
[Methods (time assignment)] Time-assignment procedure (described in the methods): the single mean time per MC move, obtained from a BD-ECMC comparison, is asserted to preserve correct rotational diffusion and collective statistics; however, the calibration regime (density, activity, packing fraction) is not specified, and no evidence is given that the mapping remains valid under steric crowding in the dense MIPS phase where effective rotational diffusion may differ from the dilute calibration.

minor comments (1)

[Abstract] Abstract: the phrase 'much faster - at comparable accuracy' would be strengthened by a brief quantitative statement of the speedup factor and the accuracy metric used.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We appreciate the referee's recognition of the potential significance of our kinetic event-chain algorithm for active matter simulations. We address each major comment below and will make revisions to improve the clarity and support for our claims.

read point-by-point responses

Referee: [Abstract] Abstract: the claim that the algorithm reproduces the phase diagram and MIPS region 'with high accuracy' is unsupported by any quantitative error metrics, binodal comparisons, or validation details; without these, the central claim cannot be evaluated.

Authors: The abstract statement regarding 'high accuracy' is indeed qualitative. The manuscript supports this through direct visual comparisons of the phase diagrams in Figure 4, where the binodals from the event-chain method closely match those from Brownian dynamics across a range of activities. To provide quantitative support, we will include error metrics, such as the maximum deviation in coexistence densities, in a revised abstract and a new table in the results section. revision: yes
Referee: [Methods (time assignment)] Time-assignment procedure (described in the methods): the single mean time per MC move, obtained from a BD-ECMC comparison, is asserted to preserve correct rotational diffusion and collective statistics; however, the calibration regime (density, activity, packing fraction) is not specified, and no evidence is given that the mapping remains valid under steric crowding in the dense MIPS phase where effective rotational diffusion may differ from the dilute calibration.

Authors: We will revise the methods section to explicitly specify the calibration conditions: comparisons were made at low densities (packing fraction φ=0.05) and various Péclet numbers in the absence of steric interactions. For the validity in the dense MIPS phase, the reproduction of the phase diagram, which includes high-density regions, serves as validation for collective statistics. However, we agree that direct evidence for rotational diffusion under crowding is lacking and will add a paragraph discussing this approximation and its implications. revision: partial

Circularity Check

0 steps flagged

No significant circularity; calibration from external BD comparison is independent validation

full rationale

The paper derives the mean time per MC move by direct comparison to Brownian dynamics (an external reference method) and then uses this to drive rotational diffusion in the event-chain algorithm. Reproduction of the phase diagram and MIPS region is presented as numerical validation against known physical results rather than a closed derivation. No equations or steps reduce the central claims to self-inputs by construction, no load-bearing self-citations are invoked for uniqueness, and the method remains falsifiable against independent BD or experimental benchmarks. This is the normal case of an algorithmic proposal with external calibration.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review provides insufficient detail to enumerate free parameters, axioms, or invented entities; the mean-time assignment appears to be a central modeling choice but its exact form and any fitted constants are not specified.

pith-pipeline@v0.9.0 · 5700 in / 1096 out tokens · 19492 ms · 2026-05-24T13:26:17.869174+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

38 extracted references · 38 canonical work pages

[1]

M. C. Marchetti, J. F. Joanny, S. Ramaswamy, T. B. Liverpool, J. Prost, M. Rao, and R. A. Simha, Rev. Mod. Phys. 85, 1143 (2013)

work page 2013
[2]

Bechinger, R

C. Bechinger, R. Di Leonardo, H. Löwen, C. Reichhardt, G. Volpe, and G. Volpe, Rev. Mod. Phys. 88, 045006 (2016)

work page 2016
[3]

Elgeti, R

J. Elgeti, R. G. Winkler, and G. Gompper, Reports Prog. Phys. 78, 056601 (2015)

work page 2015
[4]

T.Surrey, F.Nedelec, S.Leibler,andE.Karsenti,Science 292, 1167 (2001)

work page 2001
[5]

Kruse, J

K. Kruse, J. F. Joanny, F. Jülicher, J. Prost, and K. Seki- moto, Phys. Rev. Lett.92, 078101 (2004)

work page 2004
[6]

Kraikivski, R

P. Kraikivski, R. Lipowsky, and J. Kierfeld, Phys. Rev. Lett. 96, 258103 (2006)

work page 2006
[7]

Wu and A

X.-L. Wu and A. Libchaber, Phys. Rev. Lett.84, 3017 (2000)

work page 2000
[8]

Vicsek and A

T. Vicsek and A. Zafeiris, Phys. Rep.517, 71 (2012)

work page 2012
[9]

Bialké, H

J. Bialké, H. Löwen, and T. Speck, Europhys. Lett.103, 30008 (2013)

work page 2013
[10]

M. E. Cates and J. Tailleur, Annu. Rev. Condens. Matter Phys. 6, 219 (2015)

work page 2015
[11]

M. R. Shaebani, A. Wysocki, R. G. Winkler, G. Gomp- per, and H. Rieger, Nat. Rev. Phys.2, 181 (2020)

work page 2020
[12]

D.LevisandL.Berthier,Phys.Rev.E 89,062301(2014)

work page 2014
[13]

9, 5045 (2018)

J.U.Klamser, S.C.Kapfer,andW.Krauth,Nat.Comm. 9, 5045 (2018)

work page 2018
[14]

J. U. Klamser, S. C. Kapfer, and W. Krauth, J. Chem. Phys. 150, 144113 (2019)

work page 2019
[15]

R. H. Swendsen and J.-s. Wang, Phys. Rev. Lett.58, 86 (1987)

work page 1987
[16]

Wolﬀ, Phys

U. Wolﬀ, Phys. Rev. Lett.62, 361 (1989)

work page 1989
[17]

E. P. Bernard, W. Krauth, and D. B. Wilson, Phys. Rev. E 80, 056704 (2009)

work page 2009
[18]

G. Miño, T. E. Mallouk, T. Darnige, M. Hoyos, J. Dauchet, J. Dunstan, R. Soto, Y. Wang, A. Rousselet, and E. Clement, Phys. Rev. Lett.106, 048102 (2011)

work page 2011
[19]

Krauth, Front

W. Krauth, Front. Phys.9, 663457 (2021)

work page 2021
[20]

Michel, S

M. Michel, S. C. Kapfer, and W. Krauth, J. Chem. Phys. 140, 054116 (2014)

work page 2014
[21]

Michel, J

M. Michel, J. Mayer, and W. Krauth, Europhys. Lett. 112, 20003 (2015)

work page 2015
[22]

T. A. Kampmann, H.-H. Boltz, and J. Kierfeld, J. Chem. Phys. 143, 044105 (2015)

work page 2015
[23]

Nishikawa and K

Y. Nishikawa and K. Hukushima, J. Phys.: Conf. Ser. 750, 012014 (2016)

work page 2016
[24]

J.Harland, M.Michel, T.A.Kampmann,andJ.Kierfeld, Europhys. Lett. 117, 30001 (2017)

work page 2017
[25]

Klement and M

M. Klement and M. Engel, J. Chem. Phys.150, 174108 (2019)

work page 2019
[26]

T. A. Kampmann, D. Müller, L. P. Weise, C. F. Vors- mann, and J. Kierfeld, Front. Phys.9, 635886 (2021)

work page 2021
[27]

Metropolis, A

N. Metropolis, A. W. Rosenbluth, M. N. Rosenbluth, A. H. Teller, and E. Teller, J. Chem. Phys. 21, 1087 (1953)

work page 1953
[28]

See Supplemental Material at [URL will be inserted by publisher] for details on the kEC algorithm and simu- lations for the phase diagram of hard disks, additional results for a sedimenting active particle, and supporting results on passive tracer motion in an active disk bath

work page
[29]

R. Ni, M. A. Stuart, and M. Dijkstra, Nat. Commun.4, 2704 (2013)

work page 2013
[30]

Levis, J

D. Levis, J. Codina, and I. Pagonabarraga, Soft Matter 13, 8113 (2017)

work page 2017
[31]

Wysocki, R

A. Wysocki, R. G. Winkler, and G. Gompper, Europhys. Lett. 105, 48004 (2014)

work page 2014
[32]

12,113017 (2010)

L.AngelaniandR.DiLeonardo,NewJ.Phys. 12,113017 (2010)

work page 2010
[33]

Kaiser, A

A. Kaiser, A. Peshkov, A. Sokolov, B. ten Hagen, H. Löwen, and I. S. Aranson, Phys. Rev. Lett. 112, 158101 (2014)

work page 2014
[34]

S. A. Mallory, C. Valeriani, and A. Cacciuto, Phys. Rev. E 90, 032309 (2014)

work page 2014
[35]

Mokhtari, T

Z. Mokhtari, T. Aspelmeier, and A. Zippelius, Europhys. Lett. 120, 14001 (2017)

work page 2017
[36]

Knežević and H

M. Knežević and H. Stark, New J. Phys. 22, 113025 (2020)

work page 2020
[37]

Elgeti and G

J. Elgeti and G. Gompper, Europhys. Lett.101, 48003 (2013)

work page 2013
[38]

simulation time

A. P. Solon, M. E. Cates, and J. Tailleur, Eur. Phys. J. Spec. Top. 224, 1231 (2015). 6 SUPPLEMENT AL MA TERIAL In this Supplementary Information, we present (i) de- tails of the kEC algorithm, (ii) additional results for a sedimenting active particle, (iii) details of the kEC sim- ulations for the phase diagram of hard disks, and (iv) supporting results ...

work page 2015

[1] [1]

M. C. Marchetti, J. F. Joanny, S. Ramaswamy, T. B. Liverpool, J. Prost, M. Rao, and R. A. Simha, Rev. Mod. Phys. 85, 1143 (2013)

work page 2013

[2] [2]

Bechinger, R

C. Bechinger, R. Di Leonardo, H. Löwen, C. Reichhardt, G. Volpe, and G. Volpe, Rev. Mod. Phys. 88, 045006 (2016)

work page 2016

[3] [3]

Elgeti, R

J. Elgeti, R. G. Winkler, and G. Gompper, Reports Prog. Phys. 78, 056601 (2015)

work page 2015

[4] [4]

T.Surrey, F.Nedelec, S.Leibler,andE.Karsenti,Science 292, 1167 (2001)

work page 2001

[5] [5]

Kruse, J

K. Kruse, J. F. Joanny, F. Jülicher, J. Prost, and K. Seki- moto, Phys. Rev. Lett.92, 078101 (2004)

work page 2004

[6] [6]

Kraikivski, R

P. Kraikivski, R. Lipowsky, and J. Kierfeld, Phys. Rev. Lett. 96, 258103 (2006)

work page 2006

[7] [7]

Wu and A

X.-L. Wu and A. Libchaber, Phys. Rev. Lett.84, 3017 (2000)

work page 2000

[8] [8]

Vicsek and A

T. Vicsek and A. Zafeiris, Phys. Rep.517, 71 (2012)

work page 2012

[9] [9]

Bialké, H

J. Bialké, H. Löwen, and T. Speck, Europhys. Lett.103, 30008 (2013)

work page 2013

[10] [10]

M. E. Cates and J. Tailleur, Annu. Rev. Condens. Matter Phys. 6, 219 (2015)

work page 2015

[11] [11]

M. R. Shaebani, A. Wysocki, R. G. Winkler, G. Gomp- per, and H. Rieger, Nat. Rev. Phys.2, 181 (2020)

work page 2020

[12] [12]

D.LevisandL.Berthier,Phys.Rev.E 89,062301(2014)

work page 2014

[13] [13]

9, 5045 (2018)

J.U.Klamser, S.C.Kapfer,andW.Krauth,Nat.Comm. 9, 5045 (2018)

work page 2018

[14] [14]

J. U. Klamser, S. C. Kapfer, and W. Krauth, J. Chem. Phys. 150, 144113 (2019)

work page 2019

[15] [15]

R. H. Swendsen and J.-s. Wang, Phys. Rev. Lett.58, 86 (1987)

work page 1987

[16] [16]

Wolﬀ, Phys

U. Wolﬀ, Phys. Rev. Lett.62, 361 (1989)

work page 1989

[17] [17]

E. P. Bernard, W. Krauth, and D. B. Wilson, Phys. Rev. E 80, 056704 (2009)

work page 2009

[18] [18]

G. Miño, T. E. Mallouk, T. Darnige, M. Hoyos, J. Dauchet, J. Dunstan, R. Soto, Y. Wang, A. Rousselet, and E. Clement, Phys. Rev. Lett.106, 048102 (2011)

work page 2011

[19] [19]

Krauth, Front

W. Krauth, Front. Phys.9, 663457 (2021)

work page 2021

[20] [20]

Michel, S

M. Michel, S. C. Kapfer, and W. Krauth, J. Chem. Phys. 140, 054116 (2014)

work page 2014

[21] [21]

Michel, J

M. Michel, J. Mayer, and W. Krauth, Europhys. Lett. 112, 20003 (2015)

work page 2015

[22] [22]

T. A. Kampmann, H.-H. Boltz, and J. Kierfeld, J. Chem. Phys. 143, 044105 (2015)

work page 2015

[23] [23]

Nishikawa and K

Y. Nishikawa and K. Hukushima, J. Phys.: Conf. Ser. 750, 012014 (2016)

work page 2016

[24] [24]

J.Harland, M.Michel, T.A.Kampmann,andJ.Kierfeld, Europhys. Lett. 117, 30001 (2017)

work page 2017

[25] [25]

Klement and M

M. Klement and M. Engel, J. Chem. Phys.150, 174108 (2019)

work page 2019

[26] [26]

T. A. Kampmann, D. Müller, L. P. Weise, C. F. Vors- mann, and J. Kierfeld, Front. Phys.9, 635886 (2021)

work page 2021

[27] [27]

Metropolis, A

N. Metropolis, A. W. Rosenbluth, M. N. Rosenbluth, A. H. Teller, and E. Teller, J. Chem. Phys. 21, 1087 (1953)

work page 1953

[28] [28]

See Supplemental Material at [URL will be inserted by publisher] for details on the kEC algorithm and simu- lations for the phase diagram of hard disks, additional results for a sedimenting active particle, and supporting results on passive tracer motion in an active disk bath

work page

[29] [29]

R. Ni, M. A. Stuart, and M. Dijkstra, Nat. Commun.4, 2704 (2013)

work page 2013

[30] [30]

Levis, J

D. Levis, J. Codina, and I. Pagonabarraga, Soft Matter 13, 8113 (2017)

work page 2017

[31] [31]

Wysocki, R

A. Wysocki, R. G. Winkler, and G. Gompper, Europhys. Lett. 105, 48004 (2014)

work page 2014

[32] [32]

12,113017 (2010)

L.AngelaniandR.DiLeonardo,NewJ.Phys. 12,113017 (2010)

work page 2010

[33] [33]

Kaiser, A

A. Kaiser, A. Peshkov, A. Sokolov, B. ten Hagen, H. Löwen, and I. S. Aranson, Phys. Rev. Lett. 112, 158101 (2014)

work page 2014

[34] [34]

S. A. Mallory, C. Valeriani, and A. Cacciuto, Phys. Rev. E 90, 032309 (2014)

work page 2014

[35] [35]

Mokhtari, T

Z. Mokhtari, T. Aspelmeier, and A. Zippelius, Europhys. Lett. 120, 14001 (2017)

work page 2017

[36] [36]

Knežević and H

M. Knežević and H. Stark, New J. Phys. 22, 113025 (2020)

work page 2020

[37] [37]

Elgeti and G

J. Elgeti and G. Gompper, Europhys. Lett.101, 48003 (2013)

work page 2013

[38] [38]

simulation time

A. P. Solon, M. E. Cates, and J. Tailleur, Eur. Phys. J. Spec. Top. 224, 1231 (2015). 6 SUPPLEMENT AL MA TERIAL In this Supplementary Information, we present (i) de- tails of the kEC algorithm, (ii) additional results for a sedimenting active particle, (iii) details of the kEC sim- ulations for the phase diagram of hard disks, and (iv) supporting results ...

work page 2015