Sliding droplets as the chemical version for identifying the number and type of needles in a haystack

It is easy to imagine a wide range of problems where it is necessary or desirable to rapidly measure or detect a chemical. Of course, there are innumerable applications in the chemical, pharmaceutical, and biomedical arenas, but there are also situations at home, in the field, and in other locations where sophisticated measurement equipment is not available. The challenge is particularly acute when a chemical is only present at trace levels. In this issue, Yao et al. demonstrated one surprising approach to this question by integrating several tools from interfacial science and microfabrication of materials (1).

Not surprisingly, the challenge of identifying trace levels of a chemical is important for health. For example, some diseases, e.g., sepsis, are caused by small number densities of endotoxins (lipopolysaccharides), which are a major component of the cell wall of gram-negative bacteria (2). Endotoxins are toxic; therefore, as a disease mitigation strategy, the FDA regulates contamination of biomedical equipment at levels of 0.1 ng/mL or smaller, which corresponds to concentrations much smaller than 1 nM.

Thus, there are cases where it is desirable to detect chemicals and their approximate concentration rapidly, even at very low concentrations, and with methods that do not require sophisticated instruments. These two goals might seem to be in conflict: Not only is it desirable to find the chemical equivalent of a needle in a haystack, but also the number density of such “needles” is desired. When considering approaches to detect trace contaminants, it may be anticipated that concentrating the chemical is a necessary strategy for an effective detection process. So, are there novel strategies for concentrating trace materials in a sample and achieving detection and possibly even making concentration measurements? Is it possible to do so with relatively easy-to-use materials that could be used in resource-poor settings?

One known strategy for addressing at least one aspect of this detection problem is to use the interfacial (or surface) tension between two materials as a sensor for concentration, since the interfacial tension of a liquid–gas or liquid–liquid interface is sensitive to contamination. For example, it has been known since at least the nineteenth century that a “camphor” boat, which can be a small floating object such as a piece of wood, with soap or surfactant at one end, can rapidly move along the surface of water; most likely, this demonstration has been seen by many readers. The movement occurs at the interface as the clean liquid–air boundary has a higher surface tension than the liquid–air boundary covered by a surfactant, e.g., soap molecules; the interface is dragged from low to high tension, dragging the “boat” with it. There are many examples of such “Marangoni motions.” Often the motions are signatures of chemical contamination and may not be desired. In controlled settings, one can imagine that the speed of the resulting movement can provide an indication of the bulk concentration.

As an example of controlling the motion using Marangoni stresses, though not for the reason of measuring the concentration of a contaminant, Cira et al. demonstrated that adjacent liquid droplets of binary liquids could, because of evaporation-induced surface tension gradients, move in response to the vapor emitted by neighboring droplets, which also makes visible the strength of the interfacial stresses (3). Can these kinds of interfacial science approaches be used in other ways to address the measurement of concentrations of dissolved components?

The sensitivity of the interfacial tension of a liquid drop exposed to air to the concentration of a dissolved chemical motivates other measurement possibilities so long as some feature of the drop is changed. One relatively recent idea for addressing this kind of problem is a so-called wettability assay, where the movement of an aqueous drop on a substrate is monitored and can be correlated with the concentration of contaminant molecules in the solution, e.g., amphiphilic molecules in water. In the typical drop-sliding assay, a drop slides steadily over an inclined surface or perhaps moves in a stick-slip like manner. In particular, a droplet in thermodynamic equilibrium on a planar smooth horizontal surface has the same contact angle with the substrate on both sides of the drop and has a symmetric shape. When the plane is inclined, the symmetry is broken, and the angle on the “downhill” side of the drop (advancing) is larger than the angle on the uphill side (receding). An equilibrium is possible with a stationary drop, though the shape is no longer symmetric, but above a critical inclination angle, which depends on the surface tension and the difference between the advancing and receding angles (contact angle hysteresis), the droplet is mobile and slides down the plane. Typically, the drops are hydrophobic, with effective contact angles >90° with the substrate, and slide on a surface for relatively small inclination angles (<10°). Quantitative measurements of this effect characterize liquids and substrates by the effective pinning force needed to maintain a drop stationary on the surface.

The critical conditions to achieve motion in a drop-sliding assay are influenced by the roughness of the substrate and the surfactant concentration. Microfabrication and nanofabrication methods have been widely used to texture surfaces, e.g., with regular arrays of cylinders, where the three-phase contact lines (solid–liquid–gas contact) are the major sources of friction for the sliding drops. In this arrangement, the drops either rest on top of the pillars, which is referred to as a Cassie–Baxter configuration, or the liquid invades the space between the pillars, which is referred to as the Wenzel state. The latter pins the drops so the former is needed for a sliding assay; to achieve this state in the advance reported in ref. 1, the substrate is infused with silicone oil. The design of surfaces to alter the movement of such sliding drops is well studied, e.g., refs. 4 and 5, including the role of various surface patterns and the influence of the surfactants.

In one experimental arrangement using such inclined-surface mobility assays, Movafaghi et al. fabricated a rough superomniphobic (both superhydrophobic and superoleophobic) TiO2 surface with a “flower-like” nanoscale structure created by an etching process (hydrothermal synthesis) (6). Subsequent UV irradiation altered the surface chemistry and was used to pattern the surface energy of the substrate. In this way, drops in the Cassie–Baxter state, where they rested on top of the surface texture, when placed on an inclined plane, either rolled freely or were trapped. By doing experiments with solutions consisting of water–ethanol mixtures, where the mixture concentration monotonically changes the surface tension, the authors showed that the substrate designs allowed sorting droplets according to the value of surface tension (in the range 28–72 mN m⁻¹). In principle, the methodology could be used in reverse to estimate the surface tension from which the mixture composition could be estimated.

In PNAS, Yao et al. utilize their insights into physical chemistry and microfabrication to demonstrate a design strategy involving self-assembled monolayers (SAMs) on patterned substrates, whereby the detection and approximate concentration of amphiphilic molecules are possible through their effect on changes in the speed of a droplet sliding on an inclined plane (1). The specific SAM used was dimethyloctadecyl [3-(trimethoxysily)propyl] ammonium chloride or DMOAP. One can think about the development as an enhanced wettability or drop-sliding assay using a textured, amphiphilic adsorbing, substrate. Results are presented demonstrating sensitivity to amphiphilic molecules at the nanomolar level and, as in the droplet sorting discussed above, the same drop-sliding assay can be used to sense relative levels of contamination by using droplets of different samples. The authors (1) note that previous studies have reported wettability assays even for concentrations as small as 1% of the critical micelle concentration (CMC). For example, in terms of orders of magnitude, a surfactant, such as sodium dodecyl sulfate (SDS), has a CMC of about 10⁻³ M; hence, even the currently sensitive wettability assays would be inadequate at concentrations <10⁻⁵ M. Yao et al.’s method is significantly more sensitive.

In PNAS, Yao et al. utilize their insights into physical chemistry and microfabrication to demonstrate a new design strategy involving self-assembled monolayers (SAMs) on patterned substrates, whereby the detection and approximate concentration of amphiphilic molecules are possible through their effect on changes in the speed of a droplet sliding on an inclined plane.

The detection advance in ref. 1 takes advantage of the sensitivity of a feature of the interface between the SAM-covered-solid and the liquid to incorporate and concentrate contaminants, which then has a quantitative impact on a drop-sliding assay. Yao et al. have combined surface textures, in the form of cylindrical pillars placed in square arrays (typical heights 30 µm, radii 5 µm, and spacing 25 µm; Fig. 1A), with SAMs deposited on the substrate. The contact of the pillars with droplets is responsible for the friction that acts at the contact lines (a silicone oil film coats the tops of the pillars), which inhibits the motion of a drop.

Fig. 1.

Drop-sliding assay with self-assembled monolayers (SAMs) coated on a topographically patterned substrate allows sorting and chemical identification of droplets containing different types and concentrations of amphiphiles. (A) SEM image of cylindrical micropillars (radii 5 µm) on a silicon substrate. (B) Fluorescence images (Scale bar, 10 µm) showing that amphiphiles are concentrated at (C) the disordered region of the SAM at the edges of the micropillars. (D) The change in the pinning force, corresponding to a change in the inclination angle in a drop-sliding assay, for three different amphiphilic molecules. (E) Results of drop-sliding experiments where drops, containing different concentrations of the surfactant SDS (labels on Right), are pinned at different positions (labels on Left) along a 6-degree inclined substrate. The panels illustrate how the SAMs are disordered at the contact lines, which act as adsorption sites for dissolved amphiphilic molecules, and so influence the pinning force on drops in the drop-sliding assay. All panels from ref. 1.

The principle microstructural discovery of ref. 1 is that the SAMs are disordered at the edge of the pillars so that contaminants in solution, such as different kinds of amphiphiles, are adsorbed into the disordered regions of SAMs, where they influence the sliding friction of the droplets (Fig. 1 B and C). The contact lines effectively act as concentrators for the contaminants. These pinning sites are shown experimentally to allow distinguishing different molecules and concentrations (Fig. 1 D and E). In some cases, the pinning force is reduced, and in other cases, it is increased. Macroscopically, the sliding drops are incredibly sensitive to small concentrations, even at the nanomolar level, owing to the fact that the disordered regions of the SAMs can adsorb molecules and are simultaneously the major sites for friction with the drops.

One can hypothesize various reasons for the influence of the adsorbed contaminants on the change in the effective pinning force. Yao et al. present evidence that the contaminants are not changing the effective interfacial tension at the liquid–solid boundary or the length over which contact occurs. Rather, the pinning force is proportional to the difference of cosines of the advancing and receding contact angles, and the increased order or disorder of the SAMs with adsorbed molecules changes the contact angle hysteresis. Finally, the authors study different microstructures, which point to further ways to manipulate and design sensitive sensors.

Overall, the approach introduced by Yao et al., which demonstrates a DMOAP SAM-functionalized liquid-infused surface, patterned with micropillars, and sensitive to amphiphiles, yields pathways for chemical identification (with some knowledge needed of possible molecules present) and their typical concentrations, even with relatively macroscopic experiments (Fig. 1E). The disordered SAM regions at the top edges of the micropillars act to adsorb and concentrate amphiphilic molecules, which, in turn, changes the friction of a drop with a substrate. The mobility of a drop is sensitive to very small concentrations of amphiphiles, though other bulk properties of the solution are not changed.

Nevertheless, Yao et al. (1) recognize that, as presently formulated and understood, this approach is not applicable to multicomponent mixtures since it is unknown how other molecules might impact the adsorption processes in the disordered regions of the SAM-covered substrate. Also, although the simplicity of the measurement ideas allows portable sensors, the materials to do the nanofabrication and microfabrication are obviously more sophisticated. Finally, it will be interesting to explore different SAMs, their sensitivity, where disordered, to molecular features for adsorption, and perhaps other substrate topographic patterns to learn if further evolution of these ideas and discriminatory sensing is possible.