Pith. sign in

REVIEW 4 major objections 8 minor 70 references

Reframing the coupling force of adaptive resolution simulation in terms of the Liouville-type hierarchy for open systems

T0 review · 4 major / 8 minor · reviewed 2026-07-31 · grok-4.5

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A large part of the AdResS thermodynamic force is the missing average force from the open-system Liouville hierarchy, given by a one-dimensional integral.

desk verdict Clean derivation of the open-boundary average force plus a practical minimal-box recipe for AdResS thermodynamic force; the F_th≈F_av claim near AT/Δ is useful but only partly stress-tested. read the letter →

arxiv 2607.24574 v1 pith:OSJBV6SZ submitted 2026-07-27 physics.comp-ph cond-mat.stat-mechphysics.chem-ph

Julian F. Hille , Rupert Klein This is my paper · ORCID
classification physics.comp-phcond-mat.stat-mechphysics.chem-ph
keywords AdaptiveresolutionsimulationthermodynamicforceLiouville-typehierarchyopenmolecularsystemsaverageboundaryLennard-Jonesfluidinterfaceartifacts
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

Adaptive resolution simulations keep a fully interacting atomistic region open by surrounding it with transition and tracer zones, but the abrupt switching of interactions creates interface artifacts that must be cancelled by an external thermodynamic force. This paper shows that, near the atomistic boundary, a substantial part of that force is exactly the average force that particles would feel through the outer interface if interactions had not been turned off. Starting from the source terms of the Liouville-type hierarchy for open systems, the authors derive a one-dimensional integral for that average force in a simple fluid. The integral immediately implies which simulation parameters the force can and cannot depend on, which in turn licenses cheaper box geometries for computing the force and supplies interfacial validation checks that do not require a large atomistic region. Molecular-dynamics tests on the supercritical Lennard-Jones fluid at three thermodynamic states confirm that the iterated force collapses onto the integral expression near the boundary and that the proposed minimal setups reproduce the correct one-body density, Maxwellian momenta, particle-crossing statistics and average force.

What carries the argument

The average boundary force F^x_av(r_∂)=πρ(r_∂^{2}I_0-I_2) derived from the hierarchy source term Ψ_n; it supplies both the physical identification of the thermodynamic force and the parameter-independence statements used for simplification and validation.

What would settle it

Compute the iterated thermodynamic force and the integral F_av on the same grid for a supercritical Lennard-Jones fluid; if they disagree throughout the half of the Δ region nearest the AT/Δ boundary, or if particle-crossing fluctuations and the measured average force through that boundary deviate from a full-atomistic reference, the claim is false.

Watch

Extended reading notes

Core claim

Near the AT/Δ boundary the thermodynamic force of AdResS is the missing average force through the Δ/TR interface, F_th(x)=F^x_av(x_Δ/TR-x), where F^x_av is the explicit one-dimensional integral πρ(r_∂^{2}I_0-I_2) obtained by marginalising the open-system Liouville hierarchy for a simple atomistic fluid.

Load-bearing premise

The residual artifact force from excluded-volume overlap, thermostat residuals and thermodynamic mismatch must fade inside the transition region before it reaches the atomistic boundary; if it does not, the force-hierarchy identification fails.

Editorial extensions

If this is right

  • The thermodynamic force near the open-system boundary is completely fixed by density, temperature, pair potential and distance to the outer interface.
  • Force iteration can be performed in a minimal box with no atomistic region and reduced transverse dimensions, then transferred unchanged to a production run.
  • Validation reduces to interfacial observables (one-body density, Maxwellian momenta, net crossing statistics, average force) that do not require a large AT domain.
  • The same integral supplies an a-priori initial guess that already matches the converged force near the boundary.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The same integral construction should extend, with only technical changes, to multi-site molecules or multi-component mixtures once the appropriate pair correlations are supplied.
  • Because the force is independent of thermostat strength and capping radius near the boundary, one can systematically weaken or localize the thermostat without re-deriving the thermodynamic force.
  • The derived potential of mean force offers a natural starting point for time-dependent or non-equilibrium AdResS reservoirs whose target density and temperature vary slowly.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 8 minor

Summary. The manuscript reinterprets the thermodynamic (coupling) force of adaptive resolution simulations (AdResS) through the Liouville-type hierarchy for open systems. Starting from the marginalized Liouville equation, the authors identify the average force through the open system's boundary (Eq. 7), and for a simple homogeneous fluid with a planar boundary reduce it to a one-dimensional integral, F^x_av(r_∂) = πρ(r_∂²I_0 − I_2) (Eqs. 23–25), with a corresponding potential (Eq. 27). They then argue that in AdResS the thermodynamic force must reintroduce precisely this missing average force through the Δ/TR interface, and hypothesize (Eq. 30) that F_th(x) = F^x_av(x_{Δ/TR} − x) on [x_{AT/Δ}, x_fade], up to a residual F̂(x;γ,r_cap) assumed to fade within Δ. This yields parameter-dependence predictions (Tab. II), a proposed 'minimal' AdResS setup without an AT region for computing F_th, and new interface-local validation criteria (boundary density, Maxwell-Boltzmann momenta, particle-crossing statistics, average force through AT/Δ). Simulations of supercritical LJ fluids at three states with three box setups show that converged F_th from minimal/no-AT/complete boxes collapse, agree with the a priori F_av near AT/Δ, and pass the proposed validations against a full-AT reference. The analytical derivation is transparent and, in my reading, correct: ρ⁽²⁾=ρ²g(r) is exact in the bulk, the angular integration in Eq. (22) is right, and Eq. (27) is consistent with F=+dV/dr_∂. The load-bearing,

Significance. If the central identification holds, the paper provides (i) a closed one-dimensional integral (Eq. 25) for the missing average force, derived without free parameters from g(r) and V, which can serve as an a priori initial guess for the thermodynamic-force iteration; (ii) testable parameter-independence predictions (Tab. II) with practical value — notably the demonstration that a 'minimal' AdResS box with no AT region suffices to compute the thermodynamic force and reproduce correct boundary statistics; and (iii) new validation criteria tied to the reservoir-coupling terms of the hierarchy rather than bulk AT-region observables. The work also supplies reproducible code: simulation scripts in a public MRMD fork with pinned commit, plus the GhostFace and MRMDAnalysis packages with pinned commits. The conceptual link it draws — the thermodynamic force as the mean-field reintroduction of the average force through the Δ/TR interface — is a useful reframing that clarifies which artifacts the force must and must not compensate. These are solid contributions to the AdResS/open-systems literature provided the empirical identification is properly controlled.

major comments (4)
  1. [§IIIC, Eq. (30); §IVC, Fig. 6] Eq. (30), §IIIC and §IVC/Fig. 6: the central empirical claim F_th(x)=F_av(x_{Δ/TR}−x) near AT/Δ is supported by agreement between the converged iterated force and the a priori integral — but F_av (via Goldman's g(r)) was the iteration's initial guess F^0_th in Eq. (3). Near AT/Δ that guess already produces a flat density, so the correction −c∇ρ is small precisely where agreement is claimed, giving the iteration little mechanism to move away from F_av there. The decisive control — iterations started without the initial guess — is asserted in one sentence in §IVC ('converged to the same result') with no figure, no x-range, and no tolerance. This control exists (the runs were performed) and should be shown. Note that Fig. 6 does show the iteration moving substantially away from the initial guess within 1σ of Δ/TR, so the iteration is demonstrably not inert; a quantitative comparison of seed
  2. [§IVC, Figs. 5–6] Related to the above: no error norm is reported for the agreement in the region x_AT/Δ ≤ x ≤ x_fade. From Fig. 5, F_av at the AT/Δ boundary is ≲0.5 ϵ/σ versus a ~2 ϵ/σ peak near Δ/TR (L_Δ = r_cut forces F_av(x_AT/Δ)=0), so the agreement band lies where both functions are small and smoothly decaying; two such functions can agree 'very well' while hiding a residual F̂ comparable to the local force magnitude. Please report a relative residual |F_th−F_av|/|F_th| (or an L2 norm) over the claimed interval, and use it to give a quantitative estimate of x_fade, which currently is never determined from data.
  3. [§IVD, Fig. 7] §IVD (and the Conclusions): the statement that 'close to the AT/∆ boundary, the only artifact left to compensate is the missing average force' is stronger than what the validation shows. The three criteria in Fig. 7 (momentum distribution, rescaled crossing statistics, measured average force through AT/Δ against the full-AT reference) establish that the *combined* action of thermostat, capping, and F_th reproduces the correct boundary statistics, i.e. Eq. (28)/(11). They do not, by themselves, establish the decomposition Eq. (29) with F̂→0. This is genuinely independent, unseeded evidence for boundary consistency and should be credited as such — but the text should distinguish it from evidence for the identification F_th=F_av itself, which rests only on the seeded comparison in Fig. 6.
  4. [§IIID, Table II] §IIID and Table II: the parameter-independence predictions are a principal selling point, but coverage is limited to L_y,z and L_AT. Independence of γ is directly testable with the existing setup and is arguably the most consequential untested entry: the theory predicts it because F_av is momentum-independent, yet the simulations use a *global* Langevin thermostat with the rather strong γ=20τ⁻¹, and §IVD itself flags that the global thermostat may affect crossing statistics. One or two additional iteration runs at different γ (and/or r_cap) would substantially strengthen Tab. II and the claim that F̂(x;γ,r_cap) is what fades rather than something the thermostat is masking.
minor comments (8)
  1. [§IVA] §IVA: the three thermodynamic states are described as '(1) low density and low temperature, (2) high density and high temperature and (3) high density and high temperature'; state (2) should presumably read 'high density and low temperature' per Table IV.
  2. [Table IV, Appendix A] Table IV and Appendix A: densities are printed as '0.296σ³'/'0.370σ³'; the exponent should be −3 (σ⁻³) as elsewhere in the manuscript.
  3. [Table VI, §IVC, Appendix A] Table VI lists 'niter 7/10' and Appendix A states n_iter=10 with convergence at iteration 7; §IVC says the procedure was 'stopped once... sufficiently flat' at the 7th iteration. Please clarify whether iterations 8–10 were run and discarded, and state the actual numerical flatness tolerance used as the stopping criterion ('sufficiently flat' and 'prescribed tolerance' in §IIA are never quantified).
  4. [§IIIA, Appendix A] §IIIA, first paragraph: 'the conditional probability density in Eq. 7' should read 'Eq. (7)'; 'Morsali et. al' in Appendix A should be 'Morsali et al.'
  5. [Fig. 5] Fig. 5: the x-axis is shifted by x_AT/Δ while the caption text discusses distances 'greater than 1σ away from the Δ/TR interface'; since x_{Δ/TR}−x_{AT/Δ}=2.5σ, readers must mentally re-shift. Marking the 1σ-from-Δ/TR position or stating the offset explicitly would help.
  6. [Fig. 7, Appendix B] Fig. 7, second row: the rescaling of the crossing distributions for the minimal box uses the factor L1/L2=20/30 from Eq. (B19); this is stated only implicitly. Please state the applied factor and the number of snapshots/windows entering each distribution so the reader can assess statistical uncertainty of the rescaled curves.
  7. [Appendix B] Appendix B is a careful but lengthy derivation of the standard CLT rescaling of Gaussian distributions; it could be condensed substantially (the result Eq. (B19) is the only part used), improving readability without loss of rigor.
  8. [§IIID] §IIID, item on r_cap independence: the argument that g(r)=0 in the capped repulsive region renders the integral insensitive to r_cap is plausible but is an assumption about g(r) at the simulated states; a one-line reference to a measured g(r) inside r_cap (e.g., from the equilibration runs) would close it.

Circularity Check

1 steps flagged · score 2.0 of 10

Mostly non-circular: F_av is an independent integral; F_th is still the density-flattening iterate. Mild concern only that F_av was the iteration seed in the region where agreement is claimed.

  1. fitted input called prediction [§IIIC Eq. (30); §IVB–C and Fig. 6]
    "F_th(x)=F^x_av(x_Δ/TR−x), for x_AT/Δ≤x≤x_fade≤x_Δ/TR. ... Using the a priori approximation to the thermodynamic force as initial guess, we performed the iterative procedure of Eq. (3). ... In the Δ region close to the AT/Δ boundary, the thermodynamic forces are indeed in very good agreement with the a priori calculated initial guess ... We observed that, compared to the thermodynamic force iterations carried out without initial guess, the initial guess spared only one iteration and converged to the same result."

    Eq. (30) is the central identification claimed from the hierarchy. Empirically it is checked by comparing the converged iterate to the same F_av that was used as F^0_th. Near AT/Δ the seeded force already produces a flat density, so the correction −c∇ρ is small precisely where agreement is advertised; the match is therefore partly protected by the seed. The paper asserts a zero-seed control in one sentence with no figure, range, or tolerance, so that control does not fully break the dependence. This is a mild, local issue about the independence of the empirical check—not a definitional collapse of F_th into F_av.

full rationale

The load-bearing analytical chain (marginalized Liouville source term → conditional two-point density for a simple fluid → planar angular reduction → one-dimensional integral F^x_av = πρ(r_∂²I_0 − I_2)) is self-contained equilibrium statistical mechanics and does not define F_th. The thermodynamic force remains the standard iterative object F^{k+1}_th = F^k_th − c∇ρ^k; F_av enters only as an a-priori initial guess and as a post-hoc comparison target. Parameter-independence claims (no AT region, reduced L_y,z) are checked by running distinct box geometries to the same iterate and by full-AT interface diagnostics (Maxwell–Boltzmann momenta, rescaled crossing fluctuations, measured F_av through AT/Δ) that are not inputs to the force iteration. Self-citations to the authors’ prior Liouville-hierarchy and AdResS papers supply the frameworks being connected; they do not force Eq. (30). The only mild circularity-adjacent issue is empirical: agreement of converged F_th with F_av near AT/Δ is reported after seeding the iteration with F_av, in the band where F_av is already small and −c∇ρ has little room to move the force; the zero-seed control is asserted in one sentence without data. That weakens the strength of the empirical support for ˆF → 0 but does not make F_th = F_av by construction. Score 2.

Assumptions & free parameters 4 free parameters · 6 assumptions · 0 invented entities

The central identification rests on standard equilibrium liquid-state factorizations, the existing open-system Liouville hierarchy modeling assumptions, planar homogeneous geometry, and the additional working hypothesis that residual interface artifacts ˆF fade inside Δ. No new physical entity is postulated. Simulation free parameters control numerics but are not tuned to force the F_th–F_av match; that match is checked after standard density-based iteration.

free parameters (4)
  • convergence prefactor c (forcemod) = 2ε/ρ
    Weight on −∇ρ in the thermodynamic-force update (Eq. 3); set to 2ε/ρ by hand for iteration stability, not derived from the hierarchy.
  • density bin width = 0.2σ
    Controls measured ρ(x) and thus iterated F_th; authors note strong sensitivity of near-Δ/TR oscillations to bin width.
  • force application region appmax = r_AT/Δ + 4.5σ
    Hand-chosen extent (r_AT/Δ+4.5σ) where F_th is applied beyond the geometric Δ region.
  • Langevin gamma and r_cap = gamma=20 τ^{-1}, r_cap≈0.824σ
    Thermostat coupling and force-capping radius shape residual artifacts ˆF; fixed at 20 τ^{-1} and ~0.82σ without a scan, though theory claims F_th near AT/Δ is independent of them if ˆF fades.
assumptions (6)
  • domain assumption Global thermal equilibrium of the universe so that f^0_{k|i} factorizes into equilibrium one- and two-particle densities times Maxwellians (Eqs. 12–15).
    Used at the start of §IIIA to eliminate momentum dependence of F_av; required for the integral to depend only on ρ, g(r), V.
  • domain assumption Simple atomistic fluid: homogeneous isotropic ρ^{(1)}=ρ, ρ^{(2)}=ρ²g(r), pair potential V(r) only.
    §IIIA–B; excludes molecular structure, multi-site, and strong inhomogeneity the authors flag for future work.
  • domain assumption Open-system Liouville hierarchy reservoir models: influx f_in from independent reservoir single-particle density or grand-canonical form (Eqs. 9–10), and average force Ψ_n via conditional reservoir density (Eqs. 6–7).
    Taken from Delle Site & Klein; AdResS Δ/TR is identified with that reservoir without re-deriving the hierarchy.
  • domain assumption Planar AT/Δ and Δ/TR interfaces with integration domain Ω^c=[x_∂,∞)×R², so only the normal component of F_av survives.
    §IIIB; spherical AdResS geometries are mentioned but not covered by the integral reduction.
  • ad hoc to paper ˆF fades by some x_fade in Δ so F_th(x)=F^x_av(x_Δ/TR−x) on [x_AT/Δ,x_fade] (Eq. 30).
    Explicitly called an assumption in §IIIC; load-bearing for all independence claims and for the minimal-box protocol.
  • standard math Standard Liouville equation and marginalization to open subsystem; classical pair-force MD.
    Background of §IIB; not re-proved.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Reframing the coupling force of adaptive resolution simulation in terms of the Liouville-type hierarchy for open systems." pith.science (2026). https://pith.science/paper/OSJBV6SZ

@misc{pith2026260724574,
  author       = {Pith},
  title        = {Pith review of: Reframing the coupling force of adaptive resolution simulation in terms of the Liouville-type hierarchy for open systems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OSJBV6SZ}},
  note         = {Machine review of arXiv:2607.24574}
}
read the original abstract

In this work, we present a novel perspective on the coupling force employed to compensate the interface artifacts prevalent in adaptive resolution simulations (AdResS) of open many-particle systems. We show that a substantial part of this "thermodynamic force" can be framed in terms of the theoretical model of the Liouville-type hierarchy for open systems. The correspondence is made explicit for the case of a simple atomistic fluid, for which a one-dimensional integral expression is derived. This enables the analysis for dependencies of the thermodynamic force on important simulation parameters, which is taken to inspire both simplifications for the numerical calculation of the thermodynamic force and new criteria for its validation that are adequate to the interfacial nature of the problem. The theoretical claims are then verified in a simulation study of the atomistic supercritical Lennard-Jones fluid at different thermodynamic states.

Figures

Figures reproduced from arXiv: 2607.24574 by the authors.

Figure 1
Figure 1. Different setups for the simulation of an open molecular system. Full-AT: Reference setup of the open system [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Situation at the open system’s boundary with particle ⃗ [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Illustration of the integral appearing in Eq. (25), projected into the [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Proposed box setups for the calculation and validation of the thermodynamic force. AdResS (complete): AdResS [PITH_FULL_IMAGE:figures/full_fig_p011_4.png]
Figure 5
Figure 5. Figure 5: Initial guesses for the thermodynamic force iterations. Each column represents one thermodynamic state. The forces [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]
Figure 6
Figure 6. Figure 6: Results of the 7th thermodynamic force iterations at the right-hand ∆/TR interface, which is marked with a vertical gray line. Each column represents one thermodynamic state. Shown are the density profiles (first row), the thermodynamic forces (second row) and the corr…
Figure 7
Figure 7. Figure 7: Results of the validation runs at the right-hand AT/ [PITH_FULL_IMAGE:figures/full_fig_p016_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

70 extracted references · 1 canonical work pages

  1. [1]

    the thermodynamic state represented byρandT,

  2. [2]

    the fluid model given byV,

  3. [3]

    Perhaps even more interesting are the properties that the thermodynamic force close to theAT/∆boundary does not depend on under the assumption of Eq

    the distancer ∆/TR to the∆/TRinterface. Perhaps even more interesting are the properties that the thermodynamic force close to theAT/∆boundary does not depend on under the assumption of Eq. (30). Firstly, the expression in Eq. (25) was derived making use of infinite planar boundaries for the open system. This is well-reflected in the fact that, for the sl...

  4. [4]

    the box dimensionsLy andL z inyandzdirections,

  5. [5]

    the properties of theTRregion, such as its widthLTR,

  6. [6]

    the widthL ∆ of the∆region,

  7. [7]

    the widthL AT of theATregion,

  8. [8]

    the thermostat coupling parameterγ,

Show all 70 references
  1. [9]

    the capping radiusrcap. 10 E. Implications for the calculation of the thermodynamic force From the conclusions we drew in section IIID, we can derive the following simplifications for the calculation of the thermodynamic force: •Initial guess - if we know the thermodynamic sta...

  2. [10]

    boundary

    Following our treatment of this function for the combined system and reservoir in Eqs. (13) and (14) we know that this requires the particles entering into the system to have the correct one-particle densityρ(xAT/∆) =ρ. •Maxwell-Boltzmann momentum distribution at theAT/∆bounda...

  3. [11]

    Praprotnik, L

    M. Praprotnik, L. Delle Site, and K. Kremer, The Journal of Chemical Physics123, 224106 (2005)

    2005

  4. [12]

    Praprotnik, L

    M. Praprotnik, L. Delle Site, and K. Kremer, Annual Review of Physical Chemistry59, 545 (2008)

    2008

  5. [13]

    Delle Site and M

    L. Delle Site and M. Praprotnik, Physics Reports693, 1 (2017)

    2017

  6. [14]

    Delle Site, C

    L. Delle Site, C. Krekeler, J. Whittaker, A. Agarwal, R. Klein, and F. Höfling, Advanced Theory and Simulations2, 1900014 (2019)

    2019

  7. [15]

    H. Wang, C. Hartmann, C. Schütte, and L. Delle Site, Physical Review X3, 011018 (2013)

    2013

  8. [16]

    Gholami, F

    A. Gholami, F. Höfling, R. Klein, and L. Delle Site, Advanced Theory and Simulations4, 2000303 (2021)

    2021

  9. [17]

    Gholami, R

    A. Gholami, R. Klein, and L. Delle Site, Advanced Theory and Simulations5, 2100212 (2022)

    2022

  10. [18]

    Krekeler, A

    C. Krekeler, A. Agarwal, C. Junghans, M. Praprotnik, and L. Delle Site, The Journal of Chemical Physics149, 024104 (2018)

    2018

  11. [19]

    Poblete, M

    S. Poblete, M. Praprotnik, K. Kremer, and L. Delle Site, The Journal of Chemical Physics132, 114101 (2010)

    2010

  12. [20]

    Fritsch, S

    S. Fritsch, S. Poblete, C. Junghans, G. Ciccotti, L. Delle Site, and K. Kremer, Physical Review Letters108, 170602 (2012)

    2012

  13. [21]

    Delle Site and R

    L. Delle Site and R. Klein, Journal of Mathematical Physics61, 083102 (2020)

    2020

  14. [22]

    Klein and L

    R. Klein and L. Delle Site, Journal of Physics A: Mathematical and Theoretical55, 155002 (2022)

    2022

  15. [23]

    P. G. Bergmann and J. L. Lebowitz, Physical Review99, 578 (1955)

    1955

  16. [24]

    J. L. Lebowitz and P. G. Bergmann, Annals of Physics1, 1 (1957)

    1957

  17. [25]

    Agarwal, J

    A. Agarwal, J. Zhu, C. Hartmann, H. Wang, and L. Delle Site, New Journal of Physics17, 083042 (2015)

    2015

  18. [26]

    Ebrahimi Viand, F

    R. Ebrahimi Viand, F. Höfling, R. Klein, and L. Delle Site, The Journal of Chemical Physics153, 101102 (2020)

    2020

  19. [27]

    Klein, R

    R. Klein, R. Ebrahimi Viand, F. Höfling, and L. Delle Site, Advanced Theory and Simulations4, 2100071 (2021)

    2021

  20. [28]

    A. B. Poma and L. Delle Site, Physical Review Letters104, 250201 (2010)

    2010

  21. [29]

    Panahian Jand, Z

    S. Panahian Jand, Z. Nourbakhsh, and L. Delle Site, Frontiers in Chemistry10, 1072665 (2022)

    2022

  22. [30]

    Panahian Jand, T

    S. Panahian Jand, T. D. Kühne, and L. Delle Site, Advanced Theory and Simulations7, 2400833 (2024)

    2024

  23. [31]

    Gholami, R

    A. Gholami, R. Klein, and L. Delle Site, Physical Review Letters129, 230603 (2022)

    2022

  24. [32]

    Donev, J

    A. Donev, J. B. Bell, A. L. Garcia, and B. J. Alder, Multiscale Modeling & Simulation8, 871 (2010)

    2010

  25. [33]

    Delgado-Buscalioni, J

    R. Delgado-Buscalioni, J. Sablić, and M. Praprotnik, The European Physical Journal Special Topics224, 2331 (2015)

    2015

  26. [34]

    Delle Site, M

    L. Delle Site, M. Praprotnik, J. B. Bell, and R. Klein, Advanced Theory and Simulations3, 1900232 (2020)

    2020

  27. [35]

    E. R. Smith and P. E. Theodorakis, Physical Chemistry Chemical Physics26, 724 (2024)

    2024

  28. [36]

    M. Lah, N. Ntarakas, T. Potisk, P. Papež, and M. Praprotnik, The Journal of Chemical Physics162, 024103 (2025)

    2025

  29. [37]

    Djurdjevac, A

    A. Djurdjevac, A. Almgren, and J. Bell, Communications in Applied Mathematics and Computational Science20, 147 (2025)

    2025

  30. [38]

    Braun, J

    E. Braun, J. Gilmer, H. B. Mayes, D. L. Mobley, J. I. Monroe, S. Prasad, and D. M. Zuckerman, Living Journal of Computational Molecular Science1, 5957 (2019)

    2019

  31. [39]

    Shadrack Jabes and C

    B. Shadrack Jabes and C. Krekeler, Computation6, 23 (2018)

    2018

  32. [40]

    Ciccotti and L

    G. Ciccotti and L. Delle Site, Soft Matter15, 2114 (2019)

    2019

  33. [41]

    Kreis, A

    K. Kreis, A. C. Fogarty, K. Kremer, and R. Potestio, The European Physical Journal Special Topics224, 2289 (2015)

    2015

  34. [42]

    Heidari, R

    M. Heidari, R. Cortes-Huerto, D. Donadio, and R. Potestio, The European Physical Journal Special Topics225, 1505 (2016)

    2016

  35. [43]

    J. H. Peters, R. Klein, and L. Delle Site, Physical Review E94, 023309 (2016). 19 Table III. General simulation parameters. Parameter Value σ1 ϵ1 m1 rcut 2.5σ rcap 0.824 174 64σ tstep 0.002τ gamma 20.0τ −1

    2016

  36. [44]

    Thaler, M

    S. Thaler, M. Praprotnik, and J. Zavadlav, The Journal of Chemical Physics153, 164118 (2020)

    2020

  37. [45]

    Werder, J

    T. Werder, J. H. Walther, and P. Koumoutsakos, Journal of Computational Physics205, 373 (2005)

    2005

  38. [46]

    E. M. Kotsalis, J. H. Walther, and P. Koumoutsakos, Physical Review E76, 016709 (2007)

    2007

  39. [47]

    N. G. van Kampen,Stochastic Processes in Physics and Chemistry, 3rd ed., North-Holland Personal Library (Elsevier, Amsterdam, Netherlands, 2007)

    2007

  40. [48]

    Hansen and I

    J.-P. Hansen and I. R. McDonald,Theory of Simple Liquids: with Applications to Soft Matter, 4th ed. (Academic Press, Amsterdam, Netherlands, 2013)

    2013

  41. [49]

    M. E. Tuckerman,Statistical Mechanics: Theory and Molecular Simulation, 2nd ed., Oxford Graduate Texts (Oxford University Press, Oxford, United Kingdom, 2023)

    2023

  42. [50]

    Eibl and J

    S. Eibl and J. F. Hille, Multi Resolution Molecular Dynamics (2025), doi:10.5281/ZENODO.17698862

    2025 doi

  43. [51]

    Smit, The Journal of Chemical Physics96, 8639 (1992)

    B. Smit, The Journal of Chemical Physics96, 8639 (1992)

    1992

  44. [52]

    Morsali, E

    A. Morsali, E. K. Goharshadi, G. Ali Mansoori, and M. Abbaspour, Chemical Physics310, 11 (2005)

    2005

  45. [53]

    Goldman, The Journal of Physical Chemistry83, 3033 (1979)

    S. Goldman, The Journal of Physical Chemistry83, 3033 (1979)

    1979

  46. [54]

    B. M. Reible, J. F. Hille, C. Hartmann, and L. Delle Site, Physical Review Research5, 023156 (2023)

    2023

  47. [55]

    B. M. Reible, C. Hartmann, and L. Delle Site, Advances in Physics: X10, 2495151 (2025)

    2025

  48. [56]

    B. M. Reible, N. Liebreich, C. Hartmann, and L. D. Site, A fast and rigorous numerical tool to measure length-scale artifacts in molecular simulations (2025), arXiv:2511.01442 [physics.comp-ph]

    2025

  49. [57]

    Heidari, K

    M. Heidari, K. Kremer, R. Potestio, and R. Cortes-Huerto, Entropy20, 222 (2018)

    2018

  50. [58]

    Heidari, K

    M. Heidari, K. Kremer, R. Potestio, and R. Cortes-Huerto, Molecular Physics116, 3301 (2018)

    2018

  51. [59]

    Winkelmann and C

    S. Winkelmann and C. Schütte,Stochastic Dynamics in Computational Biology, Frontiers in Applied Dynamical Systems: Reviews and Tutorials (Springer International Publishing, Cham, Switzerland, 2020)

    2020

  52. [60]

    M. J. del Razo, D. Frömberg, A. V. Straube, C. Schütte, F. Höfling, and S. Winkelmann, Letters in Mathematical Physics 112, 49 (2022)

    2022

  53. [61]

    del Razo and L

    M. del Razo and L. Delle Site, SciPost Physics18, 001 (2025)

    2025

  54. [62]

    M. P. Allen and D. J. Tildesley,Computer Simulation of Liquids, 2nd ed. (Oxford University Press, Oxford, United Kingdom, 2017)

    2017

  55. [63]

    Frenkel and B

    D. Frenkel and B. Smit,Understanding Molecular Simulation, 3rd ed. (Academic Press, London, United Kingdom, 2023)

    2023

  56. [64]

    Leimkuhler and C

    B. Leimkuhler and C. Matthews,Molecular Dynamics: With Deterministic and Stochastic Numerical Methods, Interdisci- plinary Applied Mathematics, Vol. 39 (Springer International Publishing, Cham, Switzerland, 2015)

    2015

  57. [65]

    Bennett, B

    L. Bennett, B. Melchers, and B. Proppe, Curta: A General-purpose High-Performance Computer at ZEDAT, Freie Uni- versität Berlin (2020)

    2020

  58. [66]

    G. B. Arfken, H. J. Weber, and F. E. Harris,Mathematical Methods for Physicists, 7th ed. (Academic Press, Oxford, United Kingdom, 2012)

    2012

  59. [67]

    L. E. Reichl,A Modern Course in Statistical Physics, 4th ed. (WILEY-VCH, Weinheim, Germany, 2016). Appendix A: Simulation details We performed molecular dynamics simulations of a fluid with truncated, shifted Lennard-Jones potential in reduced units (σ=ϵ=m= 1). This potential ...

    2016

  60. [68]

    (23.56) on p

    Relations between normal distributions The Gaussian normal distribution with meanµand standard deviationσis defined as (see Eq. (23.56) on p. 1155 of ref. [56]) f(x;µ, σ) = 1√ 2πσ 2 exp − (x−µ) 2 2σ2 , (B1) which is related to the standard Gaussian normal distribution ϕ(x) = 1...

  61. [69]

    400 of ref

    Normal distributions from measurements From the central limit theorem (see appendix A.1.4.1 on p. 400 of ref. [57]), we know that the distribution fXN (x)of the sample meanX N = 1 N PN i=1 Yi forNindependent measurements of the identically distributed (IID) random variableY i ...

  62. [70]

    Each particle crossing from right to left is counted as+1(entering the open system) and each particle crossing from left to right is counted as −1(exiting the open system)

    Distributions of particle crossings LetC L,∆τ a,k be the net number of particle crossings through a planeaparallel to theyz-plane with areaA=L 2 within a time intervalkwith duration∆τmeasured in an equilibrium molecular fluid. Each particle crossing from right to left is count...

Pith tools

Reviewed July 31, 2026 · model on record in the stance chip above.