Intrusion and extrusion of liquids in highly confining media: bridging fundamental research to applications

ABSTRACT

Wetting and drying of pores or cavities, made by walls that attract or repel the liquid, is a ubiquitous process in nature and has many technological applications including, for example, liquid separation, chromatography, energy damping, conversion, and storage. Understanding under which conditions intrusion/extrusion takes place and how to control/tune them by chemical or physical means are currently among the main questions in the field. Historically, the theory to model intrusion/extrusion was based on the mechanics of fluids. However, the discovery of the existence of metastable states, where systems are kinetically trapped in the intruded or extruded configuration, fostered the research based on modern statistical mechanics concepts and more accurate models of the liquid, vapor, and gas phases beyond the simplest sharp interface representation. In parallel, inspired by the growing number of technological applications of intrusion/extrusion, experimental research blossomed considering systems with complex chemistry and pore topology, possessing flexible frameworks, and presenting unusual properties, such as negative volumetric compressibility. In this article, we review recent theoretical and experimental progresses, presenting it in the context of unifying framework. We illustrate also emerging technological applications of intrusion/extrusion and discuss challenges ahead.

Graphical Abstract

KEYWORDS:

• Intrusion-extrusion
• wetting-drying
• hydrophobicity
• lyophobicity
• nucleation

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. Here we used the loose notation $\mathrm{\Omega }\left({\mathrm{\Sigma }}_{LB};{\mathrm{V}}_B\right)$, while, more formally, we should have used square brackets denoting functionals rather than functions of ${\mathrm{\Sigma }}_{LB}$.

2. In the original article, the author assigned to the Tolman length a positive sign. This depends only on the sign convention of the equation for ${\gamma }_{LB}\left(R\right)$, which is opposite from that used in the present manuscript.

3. A form of $\bar{\chi }\left(r_i\right)$ converging to the characteristic function can be obtained by positioning a Gaussian function of width $\alpha$, $g_{\alpha }\left(\left(r-r_i\right)\right)$, on each fluid particle and defining $\bar{\chi }\left(r_i\right)=\underset{domain}{\overset{\hspace{0.17em}}{\int }}dr{g}_{\alpha }\left(\left(r-r_i\right)\right)$: in the limit of infinitely narrow Gaussian, $\alpha \to 0$, $\bar{\chi }\left(r_i\right)$ converges to $\chi \left(r_i\right)$. If $\alpha$ is sufficiently narrow with respect the size of the domain within which we want to count the number of fluid particles, $\underset{domain}{\overset{\hspace{0.17em}}{\int }}dr{g}_{\alpha }\left(\left(r-r_i\right)\right)$ can be re-written as the product of error functions of the distance of the particle $i$ from the border of the domain. The error function is not numerically efficient to handle on a computer and is, hence, replaced by some convenient approximation, e.g. the Fermi function $1/\left(exp\left(\lambda d\right)+1\right)$, with $d$ the distance of the particle from the border of the domain and $\lambda$ the parameter controlling the smoothness of the fermi function.

4. RMD, String, INDUS have been implemented in ‘modified’ versions of LAMMPS, GROMACS, other codes. Reader interested in running calculations of the kind described in this review and having insufficient coding experience are welcome to contact the authors to receive support.

Additional information

Funding

This project has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No. 101017858