Pith. sign in

REVIEW 2 major objections 4 minor 59 references

Accelerating equilibrium isotope effect calculations: II. Stochastic implementation of direct estimators

T0 review · 2 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read Combining stepwise direct estimators with a stochastic change of mass lets a single path integral Monte Carlo simulation evaluate large equilibrium isotope effects with zero thermodynamic integration error and lower statistical error than…

desk verdict Solid methods paper; SDE is a real and useful combination, but the low-temperature lambda-mixing caveat keeps it conditional rather than definitive. read the letter →

arxiv 1909.02910 v2 pith:GSGNQ2EW submitted 2019-08-20 physics.chem-ph physics.comp-phquant-ph

Konstantin Karandashev , Jiri Vanicek This is my paper · ORCID
classification physics.chem-phphysics.comp-phquant-ph
keywords equilibriumisotopeeffectspathintegralMonteCarlodirectestimatorsstochasticchangeofmassthermodynamicintegrationfreeenergyperturbationmethanedeuterationmethanium
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

Equilibrium isotope effects—ratios of quantum partition functions of isotopically substituted molecules—are normally computed with path integral simulations that must control three sources of error: discretization error from a finite Trotter number, statistical error from Monte Carlo sampling, and thermodynamic integration error from numerically integrating over a mass-change parameter. This paper argues that the integration error can be made exactly zero by combining two previously separate strategies: direct estimators, which evaluate a partition-function ratio by free-energy perturbation, and a stochastic change of mass, in which a single Monte Carlo trajectory visits many intermediate masses. The resulting 'stochastic direct estimators' factor a large isotope effect into a product of small factors and evaluate every factor in the same simulation, so one run yields not just the end-point ratio but the whole sequence of deuteration isotope effects. In tests on an exactly solvable harmonic model, methane, and methanium, the method matches thermodynamic integration and the stepwise direct estimators, and in short runs it reaches a converged result with fewer total potential-energy evaluations than methods requiring many separate simulations.

What carries the argument

The central object is the mass-scaled direct estimator $Z_{\mathrm{sc}}^{\lambda',\lambda''}$ of Eq. (14), whose coordinates are scaled as $r^{(s)}_{\lambda',\lambda'',i}=r^{(C)}_{i}+\sqrt{m_i(\lambda')/m_i(\lambda'')}(r^{(s)}_{i}-r^{(C)}_{i})$, so that a partition-function ratio becomes a Boltzmann-factor ratio of potential energies. The load-bearing identity is the stepwise factorization $Q_P(1)/Q_P(0)=\prod_{j=1}^{J}\langle Z_{\mathrm{sc}}^{\lambda_j,\lambda_j}\rangle^{(\lambda_j)}/\langle Z_{\mathrm{sc}}^{\lambda_j,\lambda_{j-1}}\rangle^{(\lambda_j)}$, which splits a large isotope effect into $J$ small factors, each evaluated with the sampling weight of a convenient reference mass $\bar{\lambda}_j$. The stochastic engine is a Monte Carlo move that changes the mass parameter $\lambda$ among the discrete values $\{\lambda_j\}$ using acceptance probabilities shaped by a piecewise-linear umbrella biasing potential; it is the same move that earlier removed the integration error from thermodynamic integration, now restricted to discrete jumps. The combination means every factor in the product is gathered from one trajectory, and the umbrella potential is adjusted so that the trajectory visits the intermediate masses roughly in proportion to the factors' statistical weights.

What would settle it

Run a low-temperature (e.g. 100 K) path integral Monte Carlo simulation for a floppy molecule with a small number of mass intervals, record the sequence of accepted mass values, and measure the integrated autocorrelation time of the mass index; if this autocorrelation time exceeds the autocorrelation time of the configurational coordinates, then increasing the run length will reduce the statistical error of the stochastic direct estimator more slowly than separate stepwise runs of the same total cost, contradicting the claimed single-run advantage.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that the free-energy-perturbation approach to isotope effects can be made practical for large mass changes by evaluating its stepwise factors inside a single path integral Monte Carlo simulation that stochastically changes the molecular mass among a discrete set of intermediate values. The estimator is built on the mass-scaled direct estimator $Z_{\mathrm{sc}}^{\lambda',\lambda''}$, which expresses a small partition-function ratio as an exponential of potential-energy differences at coordinates scaled by the square root of the mass ratio. When these small ratios are multiplied along the mass path, the product has zero integration error by construction, and sharing a single trajectory among all factors reduces the total computational cost and improves convergence in the nonergodic short-time regime. Numerically, the method reproduces exact finite-Trotter harmonic isotope effects, agrees with thermodynamic integration, stochastic thermodynamic integration, stepwise direct estimators, and the original direct estimators on $\mathrm{CD}_{4}/\mathrm{CH}_{4}$ and $\mathrm{CD}_{5}^{+}/\mathrm{CH}_{5}^{+}$, and extracts all sequential isotope effects $\mathrm{CH}_{4-x}\mathrm{D}_{x}/\mathrm{CH}_{4}$ and $\mathrm{CH}_{5-x}\mathrm{D}^{+}_{x}/\mathrm{CH}^{+}_{5}$ from one run. The paper also finds that the original non-stepwise mass-scaled direct estimator, previously expected to fail for larger isotope effects, remains accurate far beyond that range, and derives sufficient conditions for its convergence and for its divergence.

Load-bearing premise

The single-simulation advantage assumes that the random walk over intermediate masses mixes quickly enough at the temperature of interest, even though the smallest mass jump is fixed by the chosen discretization and the paper notes that this becomes less efficient at lower temperatures.

Editorial extensions

If this is right

  • A single stochastic run converges to a given statistical error with fewer total potential-energy evaluations than stepwise direct estimators, which must equilibrate $J$ separate simulations; the advantage is largest in the short-run, nonergodic regime.
  • All sequential isotope effects of the form $\mathrm{CH}_{4-x}\mathrm{D}_{x}/\mathrm{CH}_{4}$ and $\mathrm{CH}_{5-x}\mathrm{D}^{+}_{x}/\mathrm{CH}^{+}_{5}$ become available from the same simulation, not only the end-point $\mathrm{CD}_{4}/\mathrm{CH}_{4}$ or $\mathrm{CD}_{5}^{+}/\mathrm{CH}_{5}^{+}$ ratio.
  • Stochastic direct estimators have zero thermodynamic integration error by construction, while deterministic thermodynamic integration retains a discretization error that vanishes only as the number of mass intervals tends to infinity.
  • In the large-$J$ limit the statistical error of stochastic direct estimators approaches that of stochastic thermodynamic integration, and the deviations are already small at moderate $J$.
  • The original mass-scaled direct estimators are applicable over a much wider range of isotope effects than previously assumed, with rigorous sufficient conditions for convergence derived in an appendix.

Reading between the lines

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

  • If the discrete $\lambda$-walk mixes more slowly than the configurational space, the single-run advantage disappears; measuring the $\lambda$ autocorrelation time as a routine diagnostic would tell practitioners when SDE should be replaced by separate stepwise runs.
  • The discretization-error estimator $W_2$ of Appendix E could be combined with higher-order factorizations of the Boltzmann operator to estimate and subtract the $O(P^{-n})$ error from one run, effectively approaching the quantum limit without a sequence of Trotter numbers.
  • Because the mass-scaled coordinate transformation is a nonlocal mapping of the kind used in targeted free-energy perturbation, the same stochastic-$\lambda$ trick could be transferred to other targeted perturbations, provided the mapping can be evaluated on demand for discrete intermediate states.
  • The surprising robustness of the original direct estimators hints that in typical deuterium substitution the dangerous divergence condition $m_i(1)>2m_i(0)$ is rarely approached, so simpler estimators may remain adequate for moderate isotope effects; the safe regime is bounded by the ratio of the heavier to lighter mass.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 4 minor

Summary. The paper proposes "stochastic direct estimators" (SDE), a path integral Monte Carlo method that combines stepwise direct estimators with the authors' earlier stochastic change of mass, so that a single simulation evaluates all factors in the factored isotope-effect expression (Eq. (17)) and therefore all sequential deuteration isotope effects (e.g., CH4-xDx/CH4 for x=1..4) with zero thermodynamic integration error. The required λ- and mass-scaled moves are described in Appendix D, optimal discretization and reference-mass choices are analyzed in Appendix C, and convergence/divergence properties of the original direct estimators are analyzed in Appendices A and B. The method is validated on an eight-dimensional harmonic model, where it reproduces exact finite-P results, and on full-dimensional methane and methanium, where five independent path integral methods agree. The paper also provides a procedure for estimating the path integral discretization error of isotope effects (Appendix E).

Significance. The paper is careful and mostly convincing: it contains nontrivial analytical derivations, validates the new estimator against exact harmonic results, cross-checks five methods on two molecular systems, reports statistical errors via block averaging, and honestly identifies the central limitation in Appendix D. If the method's single-simulation advantage holds in the intended low-temperature regime, it is practically valuable because it removes thermodynamic integration error, returns all sequential isotope effects from one run, and improves convergence when computational resources are limited. However, the central advantage depends on rapid exploration of the λ dimension, and the evidence provided does not yet establish that dependence in the regime where the largest isotope effects occur.

major comments (2)
  1. [Section IIF, Eq. (17), Appendix D] The unbiasedness of Eq. (17) in a single SDE run requires the stochastic λ chain to visit every reference value λ_j sufficiently often so that each conditional average ⟨Z⟩^{(λ_j)} converges. This is a mixing requirement, not merely a variance requirement. Appendix D explicitly states that the simple λ-move 'becomes less efficient at lower temperatures' because the smallest λ step is limited by J, and Section II of the Supplementary Material reports a model in which λ exploration is slow. No quantitative diagnostic (acceptance rates, transition counts between λ values, or a lower-temperature benchmark) is provided. Since the numerical tests cover only T ≥ 200 K, the paper does not yet establish the method's central advantage in the low-temperature/large-isotope-effect regime. Please add such diagnostics or explicitly restrict the claim to the temperature range in which λ mixing is verified.
  2. [Section IIF and Figs. 2-5] The abstract and Section IIF attribute a statistical-error reduction to the stochastic λ moves via a sample-reshuffling argument, but the RMSE panels (Figs. 2(c) and 4(c)) show SDE and DE errors that are statistically indistinguishable at most temperatures. The clear, demonstrated advantage of SDE is single-run ergodicity and the ability to output all sequential isotope effects from one simulation, as illustrated in Figs. 3 and 5. Please either demonstrate a regime in which the variance is actually reduced or revise the wording so that the claimed benefit is stated as single-run ergodicity rather than decreased statistical error.
minor comments (4)
  1. [Table II (caption and header)] The table header 'ln(IE) (CD+5/CD+5)' should read 'CD+5/CH+5'; as printed, the ratio is between identical species.
  2. [Abstract vs. Reference 3] The abstract cites Cheng and Ceriotti, J. Chem. Phys. 141, 244112 (2015), while the reference list gives 2014; the two should be reconciled.
  3. [Appendix B] The phrase 'Z^{0,1}_{sc} is bound' should be 'is bounded', and likewise for 'bound observable'.
  4. [Section III.A] The phrase 'the same number of different Monte Carlo steps' should probably read 'the same number of Monte Carlo steps'.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the stochastic direct estimator is a factorization of an exact Zwanzig-type identity (Eq. 17), and the self-referential umbrella condition in Eq. (D1) is only a variance-reduction bias, not an input to the reported isotope effects.

full rationale

The central derivation is self-contained. Equations (12)-(14) obtain direct estimators from the standard Zwanzig free-energy perturbation identity, and Eq. (17) is an exact factorization of the isotope effect into conditional averages over the unbiased path-integral weights at reference masses. The stochastic change of mass is imported from the authors' earlier Ref. 16, but the final estimator in Eq. (17) remains unbiased because the discrete lambda moves sample the conditional distribution of lambda given the path, while the path distribution at each fixed lambda is the physical ring-polymer weight. The only self-referential element is the umbrella-potential update condition in Eq. (D1), which sets the bias so that the sampled lambda distribution is flattened; this condition involves the same thermal averages that appear in Eq. (17), but it is a self-consistency condition for statistical efficiency, not a fitted parameter that forces the reported isotope effects. The numerical validation is external: the harmonic model is checked against exact analytic finite-Trotter values, and the molecular results for CH4 and CH5+ are cross-checked by five independent estimators (TI, STI, DE, ODE, SDE). Self-citations to Ref. 16 provide the lambda-move machinery and a prior proof of zero integration error for STI; these are parameter-free mathematical results that do not include the target isotope effects and are not load-bearing in a circular sense. Appendix D explicitly admits that the simple lambda move becomes less efficient at low temperatures because the smallest lambda step is limited by J, an ergodicity caveat that affects convergence and correctness at low T but is not a circular reduction of the prediction to its inputs. The paper therefore contains no step in which a reported result is equivalent by construction to a fitted input or to an unverified self-citation.

Assumptions & free parameters 1 free parameters · 6 assumptions · 0 invented entities

No physical parameter is fitted to the target isotope effects. The only tunable quantities are algorithmic convergence parameters (Trotter number P, number of lambda intervals J, and the umbrella potential U_b) chosen to control statistical and discretization error; the umbrella potential is adjusted self-consistently from Eq. (D1) but does not affect unbiased averages. The central claim rests on standard path integral theory and on the numerical convergence of the Monte Carlo sampling.

free parameters (1)
  • Umbrella potential U_b(lambda_j) = self-consistently updated from simulated averages via Eq. (D1)
    A biasing potential introduced to reduce statistical error in the stochastic lambda moves; it changes sampling efficiency but cancels in the unbiased averages, so the central result does not depend on its exact form.
assumptions (6)
  • standard math Feynman path integral representation of the quantum partition function via Trotter factorization (Eqs. 3-6 in Sec. II.A).
    This is the foundational transformation used to define the ring polymer system; the paper relies on it without proof.
  • standard math Centroid virial estimator for the mass derivative dF/dlambda (Eq. 9) has path-integral-error-free statistics.
    Established in Refs. 12 and 16; the paper uses it for thermodynamic integration and stochastic thermodynamic integration reference calculations.
  • domain assumption Inverse-square-root mass interpolation (Eq. 7) is the optimal or near-optimal interpolation.
    Adopted from Ceriotti and Markland (Ref. 1) and used in all numerical examples; the paper notes the rigorous optimality only for harmonic systems in the low-temperature limit.
  • domain assumption The Jin-Braams-Bowman potential energy surface (Ref. 47) accurately describes methanium.
    The methanium isotope effects depend on this external potential energy surface; errors in it would change the numbers but not the methodological comparison.
  • standard math Block-averaging (Ref. 44) yields reliable statistical error estimates for correlated Monte Carlo samples.
    Used throughout to assign error bars to the isotope effects.
  • standard math For convex or bound potentials, the mass-scaled direct estimator Zsc has finite root mean square error (Appendix B).
    This sufficient condition underlies the explanation of why original direct estimators converge for large isotope effects.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Accelerating equilibrium isotope effect calculations: II. Stochastic implementation of direct estimators." pith.science (2026). https://pith.science/paper/GSGNQ2EW

@misc{pith2026190902910,
  author       = {Pith},
  title        = {Pith review of: Accelerating equilibrium isotope effect calculations: II. Stochastic implementation of direct estimators},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GSGNQ2EW}},
  note         = {Machine review of arXiv:1909.02910}
}
abstract

Path integral calculations of equilibrium isotope effects and isotopic fractionation are expensive due to the presence of path integral discretization errors, statistical errors, and thermodynamic integration errors. Whereas the discretization errors can be reduced by high-order factorization of the path integral and statistical errors by using centroid virial estimators, two recent papers proposed alternative ways to completely remove the thermodynamic integration errors: Cheng and Ceriotti [J. Chem. Phys. 141, 244112 (2015)] employed a variant of free-energy perturbation called "direct estimators," while Karandashev and Van\'{\i}\v{c}ek [J. Chem. Phys. 143, 194104 (2017)] combined the thermodynamic integration with a stochastic change of mass and piecewise-linear umbrella biasing potential. Here we combine the former approach with the stochastic change of mass in order to decrease its statistical errors when applied to larger isotope effects, and perform a thorough comparison of different methods by computing isotope effects first on a harmonic model, and then on methane and methanium, where we evaluate all isotope effects of the form $\mathrm{CH}_{\mathrm{4-x}}\mathrm{D}_{\mathrm{x}}/\mathrm{CH}_{4}$ and $\mathrm{CH}_{\mathrm{5-x}}\mathrm{D}^{+}_{\mathrm{x}}/\mathrm{CH}^{+}_{5}$, respectively. We discuss thoroughly the reasons for a surprising behavior of the original method of direct estimators, which performed well for a much larger range of isotope effects than what had been expected previously.

Figures

Figures reproduced from arXiv: 1909.02910 by the authors.

Figure 1
Figure 1. Isotope effect calculations in an eight-dimensional harmonic model from Ref. 16. Results of thermodynamic integration [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
Figure 3
Figure 3. Impact of nonergodicity appearing in shorter cal [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figure 5
Figure 5. Impact of nonergodicity appearing in shorter cal [PITH_FULL_IMAGE:figures/full_fig_p009_5.png] view at source ↗
Figures from the paper (3 more)
Figure 6
Figure 6. Figure 6: Relative root mean square error (RMSE) of [PITH_FULL_IMAGE:figures/full_fig_p015_6.png]
Figure 7
Figure 7. Figure 7: Relative root mean square error (RMSE) of the [PITH_FULL_IMAGE:figures/full_fig_p016_7.png]
Figure 8
Figure 8. Figure 8: Comparison of exact analytical values of the path integral discretization error of (a) the partition function [PITH_FULL_IMAGE:figures/full_fig_p018_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

59 extracted references · 41 canonical work pages

  1. [1]

    Ceriotti \ and\ author T

    author author M. Ceriotti \ and\ author T. E. \ Markland ,\ https://doi.org/10.1063/1.4772676 journal journal J. Chem. Phys. \ volume 138 ,\ eid 014112 ( year 2013 ) NoStop

    work page doi:10.1063/1.4772676 2013

  2. [2]

    Mar s \' a lek , author P.-Y

    author author O. Mar s \' a lek , author P.-Y. \ Chen , author R. Dupuis , author M. Benoit , author M. M\' e heut , author Z. Ba c i\' c , \ and\ author M. E. \ Tuckerman ,\ https://doi.org/10.1021/ct400911m journal journal J. Chem. Theory Comput. \ volume 10 ,\ eid 1440 ( year 2014 ) NoStop

    work page doi:10.1021/ct400911m

  3. [3]

    Cheng \ and\ author M

    author author B. Cheng \ and\ author M. Ceriotti ,\ https://doi.org/10.1063/1.4904293 journal journal J. Chem. Phys. \ volume 141 ,\ eid 244112 ( year 2014 ) NoStop

    doi:10.1063/1.4904293 2014

  4. [4]

    author author K. E. \ Janak \ and\ author G. Parkin ,\ https://doi.org/10.1021/ja0362611 journal journal J. Am. Chem. Soc. \ volume 125 ,\ eid 13219 ( year 2003 ) NoStop

    work page doi:10.1021/ja0362611 2003

  5. [5]

    Wolfsberg , author W

    author author M. Wolfsberg , author W. A. V. \ Hook , author P. Paneth , \ and\ author L. P. N. \ Rebelo ,\ @noop title Isotope Effects in the Chemical, Geological and Bio Sciences \ ( publisher McGraw-Hill ,\ year 2010 ) NoStop

    work page 2010

  6. [6]

    author author H. C. \ Urey ,\ https://doi.org/10.1039/JR9470000562 journal journal J. Chem. Soc. \ volume 1947 ,\ eid 562 ( year 1947 ) NoStop

    work page doi:10.1039/jr9470000562 1947

  7. [7]

    author author R. P. \ Feynman \ and\ author A. R. \ Hibbs ,\ @noop title Quantum mechanics and path integrals \ ( publisher McGraw-Hill ,\ year 1965 ) NoStop

    work page 1965

  8. [8]

    Chandler \ and\ author P

    author author D. Chandler \ and\ author P. G. \ Wolynes ,\ https://doi.org/10.1063/1.441588 journal journal J. Chem. Phys. \ volume 74 ,\ eid 4078 ( year 1981 ) NoStop

    doi:10.1063/1.441588 1981

Show all 59 references
  1. [9]

    author author D. M. \ Ceperley ,\ https://doi.org/10.1103/RevModPhys.67.279 journal journal Rev. Mod. Phys. \ volume 67 ,\ eid 279 ( year 1995 ) NoStop

    1995 doi

  2. [10]

    author author J. G. \ Kirkwood ,\ https://doi.org/10.1063/1.1749657 journal journal J. Chem. Phys. \ volume 3 ,\ eid 300 ( year 1935 ) NoStop

    1935 doi

  3. [11]

    Van\' i c ek , author W

    author author J. Van\' i c ek , author W. H. \ Miller , author J. F. \ Castillo , \ and\ author F. J. \ Aoiz ,\ https://doi.org/10.1063/1.1946740 journal journal J. Chem. Phys. \ volume 123 ,\ eid 054108 ( year 2005 ) NoStop

    2005 doi

  4. [12]

    Van\' i c ek \ and\ author W

    author author J. Van\' i c ek \ and\ author W. H. \ Miller ,\ https://doi.org/10.1063/1.2768930 journal journal J. Chem. Phys. \ volume 127 ,\ eid 114309 ( year 2007 ) NoStop

    2007 doi

  5. [13]

    Zimmermann \ and\ author J

    author author T. Zimmermann \ and\ author J. Van\' i c ek ,\ https://doi.org/10.1063/1.3167353 journal journal J. Chem. Phys. \ volume 131 ,\ eid 024111 ( year 2009 ) NoStop

    2009 doi

  6. [14]

    Zimmermann \ and\ author J

    author author T. Zimmermann \ and\ author J. Van\' i c ek ,\ https://doi.org/10.1007/s00894-010-0711-y journal journal J. Mol. Model. \ volume 16 ,\ eid 1779 ( year 2010 ) NoStop

    2010 doi

  7. [15]

    P\' e rez \ and\ author O

    author author A. P\' e rez \ and\ author O. A. \ von Lilienfeld ,\ https://doi.org/10.1021/ct2000556 journal journal J. Chem. Theory Comput. \ volume 7 ,\ eid 2358 ( year 2011 ) NoStop

    2011 doi

  8. [16]

    Karandashev \ and\ author J

    author author K. Karandashev \ and\ author J. Van\' i c ek ,\ https://doi.org/10.1063/1.4981260 journal journal J. Chem. Phys. \ volume 146 ,\ eid 184102 ( year 2017 ) NoStop

    2017 doi

  9. [17]

    Liu \ and\ author B

    author author Z. Liu \ and\ author B. J. \ Berne ,\ https://doi.org/10.1063/1.465904 journal journal J. Chem. Phys. \ volume 99 ,\ eid 6071 ( year 1993 ) NoStop

    1993 doi

  10. [18]

    Kong \ and\ author C

    author author X. Kong \ and\ author C. L. \ Brooks III ,\ https://doi.org/10.1063/1.472109 journal journal J. Chem. Phys. \ volume 105 ,\ eid 2414 ( year 1996 ) NoStop

    1996 doi

  11. [19]

    Guo , author C

    author author Z. Guo , author C. L. \ Brooks III , \ and\ author X. Kong ,\ https://doi.org/10.1021/jp972699 journal J. Phys. Chem. B \ volume 102 ,\ eid 2032 ( year 1998 ) NoStop

    1998 doi

  12. [20]

    Bitetti-Putzer , author W

    author author R. Bitetti-Putzer , author W. Yang , \ and\ author M. Karplus ,\ https://doi.org/10.1016/S0009-2614(03)01057-1 journal journal Chem. Phys. Lett. \ volume 377 ,\ eid 633 ( year 2003 ) NoStop

    2003 doi

  13. [21]

    author author R. H. \ Swendsen \ and\ author J.-S. \ Wang ,\ https://doi.org/10.1103/PhysRevLett.57.2607 journal journal Phys. Rev. Lett. \ volume 57 ,\ eid 2607 ( year 1986 ) NoStop

    1986 doi

  14. [22]

    Hukushima \ and\ author K

    author author K. Hukushima \ and\ author K. Nemoto ,\ https://doi.org/10.1143/JPSJ.65.1604 journal journal J. Phys. Soc. Jpn. \ volume 65 ,\ eid 1604 ( year 1996 ) NoStop

    1996 doi

  15. [23]

    Predescu , author M

    author author C. Predescu , author M. Predescu , \ and\ author C. V. \ Ciobanu ,\ https://doi.org/10.1021/jp045073 journal J. Phys. Chem. B \ volume 109 ,\ eid 4189 ( year 2005 ) NoStop

    2005 doi

  16. [24]

    author author R. W. \ Zwanzig ,\ https://doi.org/10.1063/1.1740409 journal journal J. Chem. Phys. \ volume 22 ,\ eid 1420 ( year 1954 ) NoStop

    1954 doi

  17. [25]

    Oostenbrink ,\ https://doi.org/10.1002/jcc.21116 journal journal J

    author author C. Oostenbrink ,\ https://doi.org/10.1002/jcc.21116 journal journal J. Comput. Chem. \ volume 30 ,\ eid 212 ( year 2009 ) NoStop

    2009 doi

  18. [26]

    Rosso , author P

    author author L. Rosso , author P. Min\' a ry , author Z. Zhu , \ and\ author M. E. \ Tuckerman ,\ https://doi.org/10.1063/1.1448491 journal journal J. Chem. Phys. \ volume 116 ,\ eid 4389 ( year 2002 ) NoStop

    2002 doi

  19. [27]

    author author J. B. \ Abrams , author L. Rosso , \ and\ author M. E. \ Tuckerman ,\ https://doi.org/10.1063/1.2232082 journal journal J. Chem. Phys. \ volume 125 ,\ eid 074115 ( year 2006 ) NoStop

    2006 doi

  20. [28]

    Wu , author X

    author author P. Wu , author X. Hu , \ and\ author W. Yang ,\ https://doi.org/10.1021/jz200808x journal journal J. Phys. Chem. Lett. \ volume 2 ,\ eid 2099 ( year 2011 ) NoStop

    2011 doi

  21. [29]

    K\" a stner \ and\ author W

    author author J. K\" a stner \ and\ author W. Thiel ,\ https://doi.org/10.1063/1.2052648 journal journal J. Chem. Phys. \ volume 123 ,\ eid 144104 ( year 2005 ) NoStop

    2005 doi

  22. [30]

    Cheng , author J

    author author B. Cheng , author J. Behler , \ and\ author M. Ceriotti ,\ https://doi.org/10.1021/acs.jpclett.6b00729 journal journal J. Phys. Chem. Lett. \ volume 7 ,\ eid 2210 ( year 2016 ) NoStop

    2016 doi

  23. [31]

    Jarzynski ,\ https://doi.org/10.1103/PhysRevE.65.046122 journal journal Phys

    author author C. Jarzynski ,\ https://doi.org/10.1103/PhysRevE.65.046122 journal journal Phys. Rev. E \ volume 65 ,\ eid 046122 ( year 2002 ) NoStop

    2002 doi

  24. [32]

    Osawa , author T

    author author T. Osawa , author T. Futakuchi , author T. Imahori , \ and\ author I.-Y. S. \ Lee ,\ https://doi.org/10.1016/j.molcata.2010.01.005 journal journal J. Mol. Catal. A \ volume 320 ,\ eid 68 ( year 2010 ) NoStop

    2010 doi

  25. [33]

    author author J. S. J. \ Hargreaves , author G. J. \ Hutchings , author R. W. \ Joyner , \ and\ author S. H. \ Taylor ,\ https://doi.org/10.1016/S0926-860X(01)00935-8 journal journal Appl. Catal. A \ volume 227 ,\ eid 191 ( year 2002 ) NoStop

    doi

  26. [34]

    Marx \ and\ author M

    author author D. Marx \ and\ author M. Parrinello ,\ https://doi.org/10.1038/375216a0 journal journal Nature \ volume 375 ,\ eid 216 ( year 1995 ) NoStop

    1995 doi

  27. [35]

    Marx \ and\ author M

    author author D. Marx \ and\ author M. Parrinello ,\ https://doi.org/10.1007/s004600050320 journal journal Z. Phys. D: At. Mol. Clusters \ volume 41 ,\ eid 253 ( year 1997 ) NoStop

    1997 doi

  28. [36]

    Marx \ and\ author M

    author author D. Marx \ and\ author M. Parrinello ,\ https://doi.org/10.1126/science.284.5411.59 journal journal Science \ volume 284 ,\ eid 59 ( year 1999 ) NoStop

    1999 doi

  29. [37]

    author author M. F. \ Herman , author E. J. \ Bruskin , \ and\ author B. J. \ Berne ,\ https://doi.org/10.1063/1.442815 journal journal J. Chem. Phys. \ volume 76 ,\ eid 5150 ( year 1982 ) NoStop

    1982 doi

  30. [38]

    Parrinello \ and\ author A

    author author M. Parrinello \ and\ author A. Rahman ,\ https://doi.org/10.1063/1.446740 journal journal J. Chem. Phys. \ volume 80 ,\ eid 860 ( year 1984 ) NoStop

    1984 doi

  31. [39]

    Van\' i c ek \ and\ author W

    author author J. Van\' i c ek \ and\ author W. H. \ Miller ,\ in\ @noop booktitle The 8th International Conference: Path Integrals from Quantum Information to Cosmology ,\ editor edited by\ editor C. Burdik , editor O. Navratil , \ and\ editor S. Posta \ ( address Prague, Czec...

    2005

  32. [40]

    author author M. E. \ Tuckerman ,\ @noop title Statistical Mechanics: Theory and Molecular Simulation \ ( publisher Oxford University Press ,\ year 2010 ) NoStop

    2010

  33. [41]

    author author W. L. \ Jorgensen \ and\ author C. J. \ Ravimohan ,\ https://doi.org/10.1063/1.449208 journal journal J. Chem. Phys. \ volume 83 ,\ eid 3050 ( year 1985 ) NoStop

    1985 doi

  34. [42]

    author author W. L. \ Jorgensen \ and\ author L. L. \ Thomas ,\ https://doi.org/10.1021/ct800011m journal journal J. Chem. Theory Comput. \ volume 4 ,\ eid 869 ( year 2008 ) NoStop

    2008 doi

  35. [43]

    author author K. S. \ Schweizer , author R. M. \ Stratt , author D. Chandler , \ and\ author P. G. \ Wolynes ,\ https://doi.org/10.1063/1.442141 journal journal J. Chem. Phys. \ volume 75 ,\ eid 1347 ( year 1981 ) NoStop

    1981 doi

  36. [44]

    Flyvbjerg \ and\ author H

    author author H. Flyvbjerg \ and\ author H. G. \ Petersen ,\ https://doi.org/10.1063/1.457480 journal journal J. Chem. Phys. \ volume 91 ,\ eid 461 ( year 1989 ) NoStop

    1989 doi

  37. [45]

    Sprik , author M

    author author M. Sprik , author M. L. \ Klein , \ and\ author D. Chandler ,\ https://doi.org/10.1103/physrevb.31.4234 journal journal Phys. Rev. B \ volume 31 ,\ eid 4234 ( year 1985 a ) NoStop

    1985 doi

  38. [46]

    Sprik , author M

    author author M. Sprik , author M. L. \ Klein , \ and\ author D. Chandler ,\ https://doi.org/10.1103/PhysRevB.32.545 journal journal Phys. Rev. B \ volume 32 ,\ eid 545 ( year 1985 b ) NoStop

    1985 doi

  39. [47]

    Jin , author B

    author author Z. Jin , author B. J. \ Braams , \ and\ author J. M. \ Bowman ,\ https://doi.org/10.1021/jp053848o journal journal J. Phys. Chem. A \ volume 110 ,\ eid 1569 ( year 2006 ) NoStop

    2006 doi

  40. [48]

    Huang , author A

    author author X. Huang , author A. B. \ McCoy , author J. M. \ Bowman , author L. M. \ Johnson , author C. Savage , author F. Dong , \ and\ author D. J. \ Nesbitt ,\ https://doi.org/10.1126/science.1121166 journal journal Science \ volume 311 ,\ eid 60 ( year 2006 ) NoStop

    doi

  41. [49]

    author author S. D. \ Ivanov , author O. Asvany , author A. Witt , author E. Hugo , author G. Mathias , author B. Redlich , author D. Marx , \ and\ author S. Schlemmer ,\ https://doi.org/10.1038/NCHEM.574 journal journal Nat. Chem. \ volume 2 ,\ eid 298 ( year 2010 ) NoStop

    doi

  42. [50]

    author author S. D. \ Ivanov , author A. Witt , \ and\ author D. Marx ,\ https://doi.org/10.1039/C3CP44523B journal journal Phys. Chem. Chem. Phys. \ volume 15 ,\ eid 10270 ( year 2013 ) NoStop

    2013 doi

  43. [51]

    Buchowiecki \ and\ author J

    author author M. Buchowiecki \ and\ author J. Van\' i c ek ,\ https://doi.org/10.1016/j.cplett.2013.09.070 journal journal Chem. Phys. Lett. \ volume 588 ,\ eid 11 ( year 2013 ) NoStop

    2013 doi

  44. [52]

    Takahashi \ and\ author M

    author author M. Takahashi \ and\ author M. Imada ,\ https://doi.org/10.1143/JPSJ.53.3765 journal journal J. Phys. Soc. Jpn. \ volume 53 ,\ eid 3765 ( year 1984 ) NoStop

    1984 doi

  45. [53]

    Suzuki ,\ https://doi.org/10.1016/0375-9601(95)00266-6 journal journal Phys

    author author M. Suzuki ,\ https://doi.org/10.1016/0375-9601(95)00266-6 journal journal Phys. Lett. A \ volume 201 ,\ eid 425 ( year 1995 ) NoStop

    1995 doi

  46. [54]

    author author S. A. \ Chin ,\ https://doi.org/10.1016/S0375-9601(97)00003-0 journal journal Phys. Lett. A \ volume 226 ,\ eid 344 ( year 1997 ) NoStop

    1997 doi

  47. [55]

    Jang , author S

    author author S. Jang , author S. Jang , \ and\ author G. A. \ Voth ,\ 10.1063/1.1410117 journal journal J. Chem. Phys. \ volume 115 ,\ eid 7832 ( year 2001 ),\ 10.1063/1.1410117 NoStop

    2001 doi

  48. [56]

    P\' e rez \ and\ author M

    author author A. P\' e rez \ and\ author M. E. \ Tuckerman ,\ https://doi.org/10.1063/1.3609120 journal journal J. Chem. Phys. \ volume 135 ,\ eid 064104 ( year 2011 ) NoStop

    2011 doi

  49. [57]

    Karandashev \ and\ author J

    author author K. Karandashev \ and\ author J. Van\' i c ek ,\ https://doi.org/10.1063/1.4935701 journal journal J. Chem. Phys. \ volume 143 ,\ eid 194104 ( year 2015 ) NoStop

    2015 doi

  50. [58]

    Azuri , author H

    author author A. Azuri , author H. Engel , author D. Doron , \ and\ author D. T. \ Major ,\ https://doi.org/10.1021/ct100716c journal journal J. Chem. Theory Comput. \ volume 7 ,\ eid 1273 ( year 2011 ) NoStop

    2011 doi

  51. [59]

    author author D. T. \ Major \ and\ author J. Gao ,\ https://doi.org/10.1016/j.jmgm.2005.04.002 journal journal J. Mol. Graph. Model. \ volume 24 ,\ eid 121 ( year 2005 ) NoStop

    2005 doi

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.