Breakdown of the Stokes-Einstein relation in Stillinger-Weber Silicon
Himani Rautela, Shiladitya Sengupta ∗, Vishwas V. Vasisht ∗ · 2025
Read Breakdown of the Stokes-Einstein relation in Stillinger-Weber Silicon by Himani Rautela, Shiladitya Sengupta ∗, Vishwas V. Vasisht ∗ (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-2503-18498.txt.
Introduction
Himani Rautela, Shiladitya Sengupta, and Vishwas Vasisht study liquid silicon in computer models in this 2025 paper. A classic rule, the Stokes-Einstein relation, links how fast atoms diffuse to how thick the liquid is. The authors test whether that rule still holds as the liquid cools below its normal freezing range.
They track diffusion, thickness, and settling times across many temperatures, pressures, and densities. The rule broke down weakly in the hot liquid when thickness was used directly. Using a settling time in its place moved the breakdown into the supercooled range. The breakdown also appeared only below a certain length scale at each temperature. The authors map these zones against other known features of silicon's phase behavior, such as where density peaks.
Catalog record
| CREATOR | Himani Rautela, Shiladitya Sengupta ∗, Vishwas V. Vasisht ∗ |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 51 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-18498
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Breakdown of the Stokes-Einstein relation in Stillinger-Weber Silicon Himani Rautela, Shiladitya Sengupta ∗, Vishwas V. Vasisht ∗ Abstract We investigate the dynamical properties of liquid and supercooled liquid silicon, modeled using the Stillinger-Weber (SW) potential, to examine the validity of the Stokes-Einstein (SE) relation. Towards this end, we examine the relationship among various dynamical quantities, including (i) the macroscopic transport coefficients - self diffusion coefficient D D and viscosity η \eta , (ii) relaxation time τ α \tau_{\alpha} as well as (iii) lengthscale dependent relaxation times τ α ( q ) \tau_{\alpha}(q) over a broad range of temperature T T , pressure P P and density ρ \rho covering both equilibrium and metastable liquid state points in the phase diagram. Our study shows a weak break down in SE relation involving D D and η \eta , and the loci of the breakdown of the SE relation (SEB) is found in the high T liquid phase. The τ α \tau_{\alpha} , when used as a proxy to η \eta , shows distinct breakdown in the SE relation whose loci is found in the supercooled liquid phase. Interestingly, certain parts of the phase diagram shows that loci of onset of slow dynamics lie below the loci of SEB, suggesting a regime that exhibits Arrhenius but non-Fickian behaviour. Computation of τ α ( q ) \tau_{\alpha}(q) enables us to extract the lengthscale associated with the Fickian to non-Fickian behaviour using which we show that the breakdown of the SE relation occurs only below a specific lengthscale at a given temperature. Further we also compare the SEB loci with other features of the phase behaviour, including the loci of compressiblity maximum, density maximum as well as diffusivity maximum. I Introduction Equilibrium liquid, when cooled faster than its typical nucleation time, remains in the liquid state even below its freezing point. This state, which is metastable relative to the crystalline phase, is referred to as a supercooled liquid. Such liquids exhibit rich physics in terms structural, dynamical and thermodynamic properties [ 1 ] . In particular, the dynamics of supercooled liquids show certain hallmark features such as the ”decoupling of time scales” - long-time structural relaxation ( \alpha -process) being characterized by multiple independent timescales, which in turn has implications on fundamental aspect of fluctuation-dissipation theorem represented in terms of the Stokes-Einstein relation (SER). The SER was originally derived for a Brownian particle of radius R , at a given temperature T , diffusing in a homogeneous fluid having a drag coefficient \zeta . The diffusion coefficient of the Brownian particle follow the relation D=\frac{k_{B}T}{\zeta} , termed as the Sutherland-Einstein-Smoluchowski equation. Using the Stokes’s law for the drag coefficient \zeta=c\pi R\eta , with c being a constant [ 2 ] and \eta being the viscosity of the background fluid, one obtains the celebrated Stokes-Einstein relation [ 3 , 4 , 5 , 6 ] given by: \displaystyle D={k_{B}T\over c\pi R\eta}\propto({\eta\over T})^{-1}. (1) Both in experiments as well as in simulation studies [ 7 ] , it has been found that Eqn. 1 remains valid at high temperature equilibrium liquids even when D is the self diffusion coefficient, up to atomic length-scales [ 8 ] . Often, with the decrease in temperature, a breakdown in this relation has been observed and ratio D\eta/T found to be temperature dependent instead of being a constant [ 9 , 10 ] . The extent of breakdown can be also expressed in term a fractional SER defined as: \displaystyle D\propto({\eta\over T})^{-\xi}. (2) where the fractional \xi exponent is 1 if the SER is valid or 0<\xi<1 in case of breakdown (SEB). The SER is typically obeyed at high temperatures, where the Fick’s law of diffusion is valid: the current \vec{j}(\vec{r},t) of local single particle density \rho(\vec{r},t) is linearly proportional to the gradient of \rho , \vec{j}=-D\nabla\rho [ 8 ] . This regime, termed as ”Fickian” regime, is characterized by a single time scale of Fickian diffusion. A high temperature liquid is also structurally homogeneous beyond single particle lengthscale [ 8 ] . However with lowering of temperature, as the liquid enters the metastable equilibrium, with growing many-body correlation the dynamics becomes spatially heterogeneous [ 11 , 12 , 13 , 14 ] . This phenomenon is known as “dynamical heterogeneity” (DH) [ 15 ] . Concurrently, various timescales of supercooled liquid dynamics get separated into two families - one proportional to the cooperative motion of the mobile particles, and the second depicting the dynamics of the immobile particles [ 16 , 17 ] . Thus, the supercooled liquid regime shows “non-Fickian” dynamics and the SEB [ 18 , 19 , 20 , 21 , 22 ] . As noted above, DH provides an intuitive physical mechanism to understand the origin of the non-Fickian nature of the dynamics [ 21 ] . However, we note that it is not the only possibility. For example, some liquids show “Fickian but non-Gaussian” dynamics [ 23 ] which may be explained by a stochastically varying “diffusing diffusivity” model [ 24 ] . Figure 1: Phase diagram in the (P,T) plane of the Stillinger-Weber silicon model . Boundaries of the stable liquid phase is shown by (i) the liquid-solid (LSC) (ii) the liquid-gas coexistence (LGC) lines along with liquid-gas critical point (light green star), and (iii) the liquid spinodal (LS) line. Also shown are the liquid-liquid coexistence (LLC) line and associated liquid-liquid critical point (blue star) deep in the supercooled regime. The region of anomaly is marked by the loci of diffusivity (TMDi), density (TMD) and compressibility maxima (TMC or the Widom line). Dotted lines along with the state points are the isobars and isochores analyzed in the present study. The loci are schematic adaptations from Refs. N. Phys. 7, 549 (2011) [ 27 ] , Copyright 2011, Springer Nature Limited and “Liquid Polymorphism: Advances in Chemical Physics”, Vol. 152 (2013), Ed. H. Eugene Stanley [ 28 ] , Copyright 2013 John Wiley & Sons, Inc. Compared to simple liquids, network-forming liquids like silicon, silica and water exhibit a richer phase behaviour. Here the inter-particle interaction favors a 4-coordinated, tetrahedrally ordered, low density local structure as opposed to a 5-coordinated, disordered local environment with higher density. Hence, on one hand, with decreasing temperature, particles tend to form an energetically stable tetrahedral network, and on the other hand, with increasing pressure (or density), the more densely packed disordered configuration is preferred [ 25 ] . This competition has two important consequences: (i) network-forming liquids can display a novel liquid-liquid phase transition and associated liquid-liquid critical point [ 26 , 27 , 28 ] and (ii) they can exhibit anomalies in various structural, dynamic and thermodynamic properties [ 29 , 30 , 31 , 32 , 33 , 34 ] . Fig. 1 illustrates the computed P-T phase diagram for silicon [ 27 , 33 ] . It shows (i) “LSC”, the liquid-solid and (ii) “LGC” the liquid-gas co-existence lines and the liquid-gas critical point; (iii) “LLC”, the liquid-liquid co-existence line ending in a critical point, existing deep inside the supercooled liquid regime; (iv) “LS”, the liquid spinodal beyond which the liquid is mechanically unstable and (v) loci of maxima of dynamic, thermodynamic and