The ability to control the pressure at which water wets (intrusion) a hydrophobic nanoporous material and the one at which it evacuates it (extrusion) is important when choosing a material for an application of heterogenous lyophobic systems (HLS), be it liquid chromatography, energy storage or the dissipation of vibrations. For these applications, the precise control of intrusion and extrusion pressure and the hysteresis that is sometimes observed has important technological implications. Some parameters that are normally considered in the design of hydrophobic nanoporous materials include the shape of the nanopores (the radius of the aperture, the internal surface area) and the surface chemistry (how hydrophobic it is);  even though these parameters offer a good prediction of, mainly, the intrusion pressure they do not tell the whole story. What role does the connection between pores and, in particular, the shape of secondary cavities have in the wetting proprieties of a nanoporous material? What can we learn from taking them into account and how can their presence be used to improve the target properties of a material?

Connections matter

Nanoporous materials, and in particular hydrophobic materials, come in a wide array of sizes and pore shapes as channels, cages, and everything in between, but this diversity in how pores are connected is seldom considered. Some zeolites, a type of material particularly used in intrusion/extrusion applications, can have very similar chemical composition and pores of relatively similar sizes, have very different intrusion and extrusion pressures. Our simulations show that materials in which the main nanopores are independent may have a radically different behavior than those connected by smaller secondary  channels. Actually, having connections between the pores can turn the overall behavior of the nanoporous materials all the way from water-loving (hydrophilic) to exceptionally water-hating (superhydrophobic).

Length of secondary channels matters

There exists, since the 1800’s, an equation that can be used to predict the pressure required for intrusion of a liquid in a pore that resists wetting (lyophobic). The Kelvin-Laplace equation is surprisingly simple, considering only the inverse of the pore diameter, the ma

terial’s contact angle (the angle that a droplet of the intruding liquid does on the surface) and the liquid surface tension. The intrusion of water in a hydrophobic nanopore can require the liquid pressure to be increased to hundreds of atmospheres, such that any pores or cavities at this scale are normally considered to be dry under ambient pressure. The fact that these pores remain dry, even when water could have access to them, effectively increases the hydrophobicity of a material, so called the lotus effect, and so one could think that the presence of subnanometric pores and cavities in the lining of a main pore could only hinder its wetting proprieties.

This recent work shows that not only can subnanometric pores get wet, but they can decrease the intrusion pressure (and increase the extrusion pressure) in a way that can be controlled. Shorter pores stay wet more often effectively serving as hydrophilic patches on the hydrophobic surface of the main pore, having an anti-lotus effect. The longer secondary channel, the lower the probability that they get wet at ambient pressure, ultimately creating a superhydrophobic materials. If the length of secondary channels can be controlled during the fabrication process, like in the case of some mesoporous silica, the intrusion and extrusion pressure can be tuned to better suit the applications one has in mind.

It’s all about hydrogen bonds

Even though the Kelvin-Laplace equation would predict an intrusion pressure of ca. 90 MPa for the secondary channels these get wet at ambient pressure. The Kelvin-Laplace equation also does not predict that different channel lengths would have different probabilities to stay wet. This effect seems to have a molecular origin: this work shows that, indeed, hydrogen-bonding between water molecules is fundamental to explain it. Water molecules that are close to the dry secondary channels have less hydrogen bonds than the molecules closer to the centre of the main pore. This means that having a water molecule confined in the secondary channel can sometimes be energetically favourable, because it allows to form a bridge of hydrogen bonds through the secondary channel. For short channels the budget is favorable, as having one or two molecules inside the channel means that both those and the ones close to the secondary channel mouths are creating more hydrogen bonds than if there was no water in the channels. For longer channels this is not the case as the cost of pushing water molecules in the middle of the secondary is higher than the gain.  

Simulations and designing nanoporous materials

The fact that subnanometric channels affect the wetting proprieties of a nanoporous materials and the role that hydrogen bonds have on allowing these channels to get wet are molecular processes that would not be captured by continuous arguments. Molecular dynamics simulations allow for the exploration of these minute phenomena while precisely tuning some parameter, in this case the secondary channel length, and measuring its effect on a macroscopic measurement, the intrusion/extrusion pressure. Not only that, having access to molecular detail allows us to think of other systems where the cooperative effect of hydrogen bonding could also be a determining factor on the wetting properties of nanoporous materials, as could be the case in the wetting of cages (not channels) in metal-organic frameworks (MOF), a family of materials that has been used in multiple intrusion/extrusion applications. Simulations can help guide the design process of nanoporous materials by laying the groundwork for understanding the molecular processes that are involved in the wetting/drying of the material.