Entropy exchange in an inter-correlating binary quasi-classical system: Concept of entropy-bath accelerated molecular dynamics
Projesh Kumar Roy · 2025
Read Entropy exchange in an inter-correlating binary quasi-classical system: Concept of entropy-bath accelerated molecular dynamics by Projesh Kumar Roy (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-15474.txt.
Introduction
Roy explores in twenty twenty-five how two linked systems can trade entropy, the measure of disorder, in a way that bends classic rules. He studies a pair made of one plain system and one quantum-like system, linked by a non-standard statistical tie. Surprisingly, the quantum-like traits appear even though that system has no self-link of its own.
From this idea he proposes a new way to speed up molecular simulations. A virtual entropy bath trades disorder with the simulated system, reshaping its energy spread. That reshaping lets stuck systems explore their energy landscape far more fully at balance. He sketches the method's steps and tests the idea on cold glass-like silica, a classic hard case.
Catalog record
| CREATOR | Projesh Kumar Roy |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 26 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-15474
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Entropy exchange in an inter-correlating binary quasi-classical system: Concept of entropy-bath accelerated molecular dynamics Projesh Kumar Roy Abstract This letter highlights the entropy exchange phenomenon in a coupled binary inter-correlating system following Haldane’s non-linear statistical correlation. A unique coupling between a classical and a quantum-like system at the marginal distribution is observed. It is shown that the quantum nature of a system can arise without any self-correlation. Extending this idea, an enhanced sampling method in molecular dynamics simulation is postulated where a classical system is forced to show quantum-like behavior with the help of an entropy-bath . An entropy-bath exchanges entropy with the system to scale the potential energy distribution of the system, so that a probability upper bound at each energy level is maintained. An algorithm to implement the entropy-bath accelerated molecular dynamics simulation is discussed. Using low temperature vitreous silica as an example, the capability of such an algorithm to greatly improve sampling of the potential energy landscape under equilibrium conditions for kinetically arrested systems is highlighted. I Introduction Intermediate statistics can be described as a unified statistical theory that encompasses Fermi-Dirac (FD), Bose-Einstein (BE), and Maxwell-Boltzmann (MB) statistics. Real-world applications of intermediate statistics have been actively explored in recent times. Among several variants of intermediate statistics [ 1 , 2 , 3 , 4 , 5 ] , Haldane-Wu (HW) statistics [ 6 , 7 ] , and its updated version Polychonakos statistics [ 8 , 9 ] , have shown promising applications in quantum [ 10 , 11 , 12 , 13 ] as well as quasi-classical systems [ 14 , 15 , 16 , 17 , 18 , 19 ] . The concept of HW statistics is based on the idea of Haldane’s statistical correlation theory [ 6 ] in quantum systems, which was extended and applied to BE statistics. Recently, I have shown [ 20 ] that a non-linear extension of Haldane’s theory can be applied to classical MB statistics to derive a novel quasi-classical intermediate statistics for a pure self-correlating system. It has a similar form as the classical fractional exclusion statistics [ 21 ] (CFES), which was derived earlier for a classical system without any quantum approximation. In this letter, I show that self-correlation is not necessary to show quantum-like characteristics in this quasi-classical model. Using a binary inter-correlating system, I prove that due to inter-correlation, one part of the system can show classical distribution, whereas the other part can show quantum-like exclusion characteristics. This unique coupling paves the way to conceptualize an entropy-bath in molecular dynamics (MD) simulation, where a classical system is forced to behave as a quantum-like system obeying the exclusion principle by exchanging entropy with a hypothetical reservoir. The letter is organized as follows: In section II II , I derive the coupled energy distributions for an inter-correlated binary system, and I explain the entropy exchange between the coupled systems. Following this, I propose a possible application of the binary system in MD simulation as an entropy-bath for a faster exploration of the potential energy landscape (PEL). In section III III , I show the effects of entropy-bath accelerated MD (EBMD) in glassy Silica. Finally, I conclude in section IV IV with an outlook. II Theory and Derivation Inter-Correlating Binary System : In all the following sections of this letter, the word particle will be used to define the building blocks of a system, whose energy distribution follows the maximum entropy (MaxEnt) principle. I consider a closed system consisting of two species; Q and S ; where neither component has any self-correlation, but they are inter-correlated to each other. Particles belonging to both Q and S components are distinguishable objects and follow MB statistics, i.e., their probability distributions ( p ) follow p=g\exp\{-\mu-\beta\epsilon\} ; where \beta=(1/k_{B}T) with k_{B}= Boltzmann constant and T= temperature of the system, \epsilon is the energy of the system with degeneracy g , and \mu is a constant related to chemical potential in thermal ensembles. The components contain a total of N=N_{Q}+N_{S} particles with N\mu=N_{S}\mu_{S}+N_{Q}\mu_{Q} , and total energy E=N_{Q}E_{Q}+N_{S}E_{S} , respectively. Let’s imagine that, at the marginal distribution, the population n_{q} at energy level \epsilon_{q} of component Q affects the population n_{s} at energy level \epsilon_{s} of component S . I describe the nature of this correlation using Haldane’s non-linear statistical correlation theory [ 20 ] , \tilde{g}_{s}=g_{s}-\sum_{q}\gamma_{sq}\Bigg(\frac{n_{q}^{m}}{g_{q}^{m-1}}\Bigg) (1) \tilde{g}_{s} and g_{s} are the available and true degenerate sates of S component at energy level \epsilon_{s} , respectively. m is an integer [ 20 ] with range 1\leq m\leq\infty . The correlation function, \gamma_{sq} , can be a function of the properties of both S and Q components. The total number of microstates ( W ) for such a coupled binary quasi-classical system can be written as, W=W^{\text{MB}}_{S}W^{\text{MB}}_{Q}\prod_{s}\Bigg\{1-\frac{1}{g_{s}}\sum_{q}\gamma_{sq}\Bigg(\frac{\nu_{q}^{m}}{g_{q}^{m-1}}\Bigg)\Bigg\}^{\nu_{s}} (2) \nu_{s} and \nu_{q} are un-optimized populations and W^{\text{MB}}_{S}=g_{s}^{\nu_{s}}/\nu_{s}! , W^{\text{MB}}_{Q}=g_{q}^{\nu_{q}}/\nu_{q}! for S and Q components, respectively. Next, the MaxEnt method is used to obtain the optimal distribution of the combined system under NVE conditions, where the E and \mu are constant. Using the MaxEnt method, we can obtain the partial derivatives of W of this system, (\partial\ln W/\partial\nu_{s}) and (\partial\ln W/\partial\nu_{q}) to maximize W and obtain the optimal populations for the S and Q components for the combined system as \nu_{q,\text{max}}\equiv n_{q} and \nu_{s,\text{max}}\equiv n_{s} , which can be mathematically expressed as, n_{s}=g_{s}A_{s}\Bigg\{1-\frac{1}{g_{s}}\sum_{q}\gamma_{sq}\Bigg(\frac{n_{q}^{m}}{g_{q}^{m-1}}\Bigg)\Bigg\} (3) n_{q}=g_{q}A_{q}\exp\Bigg\{-m\Bigg(\frac{n_{q}^{m-1}}{g_{q}^{m-1}}\Bigg)\sum_{s}\Bigg(\frac{\gamma_{sq}n_{s}}{\tilde{g}_{s}}\Bigg)\Bigg\} (4) where A_{s}=\exp(N_{S}\mu_{S}-\beta\epsilon_{s}) , A_{q}=\exp(N_{Q}\mu_{Q}-\beta\epsilon_{q}) . To separate s - and q -dependent terms in the summand in Eq. 4 , I make an approximation that \gamma_{sq} is separable as, \gamma_{sq}\equiv\gamma_{s}\gamma_{q} . For dilute systems; i.e., g_{s}>>n_{s} and g_{q}>>n_{q} ; the final equation for the energy distribution of the Q component reads, n_{q}=g_{q}A_{q}\Bigg\{1-m\langle\zeta_{S}\rangle\gamma_{q}\Bigg(\frac{n_{q}^{m-1}}{g_{q}^{m-1}}\Bigg)\Bigg\} (5) \langle\zeta_{S}\rangle is effectively an average of the \gamma_{s}/\tilde{g}_{s} factor. At the marginal distribution where the entropy of both components is maximized, I can substitute the values of Eq. 3 in Eq. 4 . Thus, for classical systems with continuous energy levels, \langle\zeta_{S}\rangle can be written as, \langle\zeta_{S}\rangle(\mu_{S},\beta)=\sum_{s}\frac{n_{s}\gamma_{s}}{\tilde{g}_{s}}\approx\int_{0}^{\infty}\gamma_{s}A_{s}\partial\epsilon_{s} (6) Thus, \langle\zeta_{S}\rangle is a function of the temperature and \mu of the S component only and therefore a constant for thermal ensembles. \langle\zeta_{S}\rangle must have a real and finite value for a realistic n_{q} distribution. It follows that \gamma_{s}\exp(-\beta\epsilon_{s}) must converge at \epsilon_{s}\to\infty following Eq. 6 . Interestingly, Eq. 5 takes the form of a CFES-type equation [ 21 , 20 ] for m\geq 2 , where \langle\zeta_{S}\rangle can be considered as the third Lagrange parameter and m being the exponent as described in reference 21 . As a result, there is an upper bound ( n^{*}_{q} ) to the particle distribution at each degenerate state g_{q} for \gamma_{s}\gamma_{q}>0 as, n^{*}_{q}=\Bigg(\frac{1}{m\langle\zeta_{S}\rangle\gamma_{q}}\Bigg)^{\Big(\frac{1}{m-1}\Big)} (7) Unlike traditional quantum systems, n^{*}_{q} is not a universal constant and may be a function of thermodynamic variables such as temperature and pressure due to the presence of \langle\zeta_{S}\rangle in Eq. 7 . In retrospect [ 21 , 20 ] , as \langle\zeta_{S}\rangle is constant at the optimal distribution, the n_{q} -distribution has an overall constraint C_{Q} as, C_{Q}=\sum_{q}\gamma_{q}\Bigg(\frac{n_{q}^{m}}{g_{q}^{m-1}}\Bigg) (8) Interestingly, the exponent in Eq. 5 and 8 is same as the purely classical CFES system described in reference 21 , and not shifted by 1 as was the case for the self-correlating quasi-classical system