Open questions on defining and computing the vapour-liquid surface tension by virial and test transformation approaches

Martin Thomas Horsch · 2025

Read Open questions on defining and computing the vapour-liquid surface tension by virial and test transformation approaches by Martin Thomas Horsch (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-2501-12755.txt.

Introduction

Horsch raises four unresolved problems in how surface tension at vapor-liquid boundaries is defined and computed, and argues each one still lacks an answer. The first asks why a kinetic term seems to appear in the tension at all. The second concerns the local virial, which some calculations seem to require but no one can define cleanly.

The third problem is practical: for curved, spherical interfaces, the virial route and the test-area route give different numbers. The fourth is the deepest. Unless values from these methods can be tied back to the thermodynamic meaning of surface tension, computing them may be pointless. Horsch leaves all four questions open, deliberately.

Catalog record

CREATORMartin Thomas Horsch
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE87 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2501-12755

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Open questions on defining and computing the vapour-liquid surface tension by virial and test transformation approaches Martin Thomas Horsch Abstract This work addresses four problems in defining and computing the surface tension of vapour-liquid interfaces: (1) The apparent kinetic contribution to the surface tension, and what it is that makes it appear; (2) the problem of defining the local virial, which for some purposes appears to be necessary; (3) the disagreement between results for spherical interfaces when using different methods, specifically the virial and the test area method; (4) how the surface tension values obtained from these methods can be related to the definition of the surface tension from thermodynamics – since if they cannot, it is a meaningless exercise to compute them. At the end, all these problems remain unsolved: They are “open questions.” 1 Introduction 1.1 Interfacial properties: Open questions Molecular modelling and simulation is often applied to the thermodynamic properties of fluid systems. Most commonly, the underlying models are defined by pairwise additive force fields which describe the intermolecular interactions by effective pair potentials. While thermodynamic properties, including for fluid phase equilibria, can be obtained by molecular simulation of homogeneous systems, also heterogeneous systems can be simulated with the same intermolecular pair potentials. In this way, in particular, interfacial properties can be computed, the most important of which is the surface tension. While this is regularly done and can be considered an established practice, fundamental open questions remain that threaten to undercut what has been claimed to be the foundation for commonly employed methods. This in turn puts much of the results from the literature into question. Here we will be considering a few such open questions with a focus on methods for computing the surface tension that are based on the virial. The local virial tensor (and pressure tensor) is always an underspecified quantity, but in the planar case, this does not matter because the volume integral over it yields a unique, well-defined result both for the overall pressure tensor and the surface tension. When integrating over the (virial as a function of the) radius in polar coordinates, this property disappears, and the surface tension is no longer well-defined. Unfortunately, results for the surface tension of droplets and bubbles from the virial route are known to disagree both with results from test transformations (test area method) and with what is known about the Tolman length for the respective systems. In particular, the surface tension of Lennard-Jones nanodroplets is significantly smaller if computed by the virial route than by the test area method [ 1 ] ; other thermodynamic methods tend to agree with the test area results [ 2 ] . Such a disagreement was found in the case of curved interfaces only, whereas for planar interfaces, the two approaches are exactly equivalent [ 3 , 4 ] . Lau et al. [ 5 ] uncovered a possible source of this disagreement by proving that for spherical interfaces, the response of the interfacial free energy to a small distortion (test transformation) contains a non-zero contribution from the variance of the response of the energy, i.e. , from a second-order contribution in terms of the internal energy of the system. As usual in molecular simulation, these methods are based on statistical mechanical expressions that are connected to phenomenological thermodynamics through the partition function and the thermodynamic potential. In the spherical case, however, this connection looks more shaky that it usually does when connecting thermodynamics to statistical mechanics. Moreover, even in the planar case, Sega et al. [ 6 ] argue that a kinetic contribution appears, and that this contribution would need to be added to that from the virial, which to date has not become common practice at all. Is this an indication that almost all molecular simulation data on the surface tension are wrong? In the case of water, for instance, the error due to neglecting the effect from the apparent kinetic contribution would amount to 15%, Sega et al. [ 6 ] find. These problems are summarized below; none of them will be resolved, instead, the aim here is to highlight their importance and provide some input to further discussions. This is proposed in the hope that the discussion also represents some actual progress on the issues mentioned. The apparent kinetic contribution to the surface tension of planar interfaces is discussed in Section 2 . Then the spherical case is discussed: The methodology and the underlying statistical mechanics in Section 3 , the perspective from phenomenological thermodynamics in Section 4 . However, after the first thermodynamic analysis, some problems remain. Concluding this work, an orientation for future efforts is proposed in Section 5 . This will all be preceded by a clarification of the scope of this work and the terminology to be used (Section 1.2 ). 1.2 Key concepts and scope of this work This section is meant as a terminological clarification. Key quantities discussed in this work are the virial and the surface tension . A term that may cause misunderstandings is “second-order virial,” a concept introduced as central to the methodology presented in Section 3 . The word virial is taken from Latin and means related to forces . In thermodynamics, it is used for the virial expansion (which this work is not about), the virial coefficients which are the coefficients that occur in the virial expansion (this work is therefore not about these either), the quantity called the virial , and its representation as a tensorial quantity, the virial tensor (the latter two are what this work is about). These all have in common that they describe how the behaviour of a fluid deviates from that of an ideal gas due to the forces acting between molecules. The virial expansion, also called the virial equation of state, is a caloric equation of state \frac{\mathnormal{P}}{\mathnormal{T}}=\mathnormal{\rho}+\mathnormal{B}_{2}(\mathnormal{T})\mathnormal{\rho}^{2}+\mathnormal{B}_{3}(\mathnormal{T})\mathnormal{\rho}^{3}+\mathnormal{B}_{4}(\mathnormal{T})\mathnormal{\rho}^{4}+\dots, (1) when expanded in terms of density, where \mathnormal{B}_{\mathnormal{k}} is called the \mathnormal{k} -th virial coefficient. Thereof, the most important one is \mathnormal{B}_{2} , the second virial coefficient , which controls the leading contribution to the deviation between ideal and real fluid behaviour. If Eq. ( 1 ) is truncated after the second term, \mathnormal{P}/\mathnormal{T}=\mathnormal{\rho}+\mathnormal{B}_{2}\mathnormal{\rho}^{2} , it becomes an expansion to the second order, which is consequently referred to as a second-order virial expansion or second-order virial equation of state . Both of these terms contain the phrase “second-order virial,” and they indeed have some relevance to the theory of nucleation and the surface tension of nanodroplets. For example, in Laaksonen-Ford-Kulmala nucleation theory [ 7 ] , the free energy of formation for clusters in a supersaturated vapour is determined from the second-order virial expansion; the key parameter required for this is \mathnormal{B}_{2} . The free energy of formation of the clusters can be converted to an estimate for their surface tension [ 8 ] . Urrutia [ 9 , 10 ] discusses the curvature dependence of the surface tension in terms of an expansion for the grand potential based on Mayer cluster integrals [ 11 ] , which can be truncated after its second term; for this, he uses the term “second-order virial series,” including in the title of one of his articles [ 10 ]

Read the complete text