Abstract
Computer simulations have been employed in recent years to evaluate the configurational entropy changes in model glass-forming liquids. We consider two methods, both of which involve the calculation of the 'intra-basin' entropy as a means for obtaining the configurational entropy. The first method involves the evaluation of the intra-basin entropy from the vibrational frequencies of inherent structures, by making a harmonic approximation of the local potential energy topography. The second method employs simulations that confine the liquid within a localized region of configuration space by the imposition of constraints; apart from the choice of the constraints, no further assumptions are made. We compare the configurational entropies estimated for a model liquid (binary mixture of particles interacting via the Lennard-Jones potential) for a range of temperatures, at fixed density.
Introduction: The Glass Transition Puzzle
When a liquid is cooled rapidly, it can bypass crystallization and become a supercooled liquid, eventually hardening into a glass. A central question in condensed matter physics is whether this transition is purely kinetic—the liquid simply becomes too viscous to move—or if there is an underlying thermodynamic phase transition.
Configurational entropy ($S_c$) is central to this puzzle, representing the number of distinct "basins" of attraction in the energy landscape. As a liquid cools, $S_c$ decreases; if it reaches zero at a finite temperature, it signals an "ideal" glass transition. However, calculating $S_c$ is notoriously difficult because it requires separating the vibrational entropy of particles jiggling within a local arrangement from the entropy of the arrangements themselves.
Methodology: Two Paths to Configurational Entropy
This study evaluates two distinct computational strategies for isolating $S_c$ in a model glass-forming liquid (a binary Lennard-Jones mixture):
- Inherent Structure (IS) Approach: This method maps liquid configurations to their nearest local potential energy minima (inherent structures). The entropy of the surrounding basin is calculated using a harmonic approximation, assuming the energy landscape is locally parabolic. $S_c$ is then obtained by subtracting this vibrational entropy from the total entropy of the liquid.
- Constrained System Simulations: This method physically traps particles within Voronoi "cages" defined by their neighbors. Monte Carlo simulations are performed where particles are forbidden from leaving these cages. Using thermodynamic integration from a reference Einstein crystal, the free energy of these trapped states is calculated directly without making assumptions about the shape of the energy landscape.
Results: The Harmonic Approximation and Deceptive Linearity
A key finding is that the harmonic approximation, while a standard tool in the field, is only strictly valid at low temperatures ($T < 0.8$). At higher temperatures, the basins in the energy landscape deviate from simple parabolas, and the approximation begins to fail.
Interestingly, the difference between the liquid's total energy and its inherent structure energy ($E - E_{IS}$) appears linear with temperature even up to high values (e.g., $T = 1.5$). This linearity can be deceptive, as it might lead researchers to assume harmonic behavior where it does not exist, as shown in the figure below.
Comparison of Entropy Estimates and the Ideal Glass Transition
The study compares the configurational entropy derived from both methods. While there is moderate agreement, the constrained simulation method predicts a slightly weaker temperature dependence for $S_c$ compared to the inherent structure method.
By extrapolating the configurational entropy data, the ideal glass transition temperature (Kauzmann temperature, $T_K$) is estimated to be $T \approx 0.294$. This result aligns well with previous independent estimates and supports the thermodynamic view of the glass transition.
Conclusion
The comparison reveals that while the inherent structure approach with a harmonic approximation is computationally efficient, it has significant limitations at higher temperatures. Constrained simulations using Voronoi cages provide a more robust, though demanding, alternative for measuring intra-basin entropy directly. Future work will focus on refining these constraints and exploring density dependence to further validate these thermodynamic models.
Summary updated