Response to Comment on “Vapor Lubrication for Reducing Water and Ice Adhesion on Poly(dimethylsiloxane) Brushes”: Organic Vapors Influence Water Contact Angles on Hydrophobic Surfaces

Fast removal of water drops from solid surfaces is important in many applications such as on solar panels in rain, in heat transfer, and for water collection. Recently, a reduction in lateral adhesion of water drops on poly(dimethylsiloxane) (PDMS) brush surfaces after exposure to various organic vapors was reported. It was attributed to the physisorption of vapor and swelling of the PDMS brushes. However, it was later pointed out that a change in the interfacial energies by vapor adsorption could also have caused low drop adhesion. To find out how strongly each effect contributes, contact angles of water drops on three hydrophobic surfaces in different vapors are measured. In water‐soluble vapors, a substantial decrease is observed in contact angles. This decrease can indeed be explained by a vapor‐induced change in the interfacial tensions. The very low contact angle hysteresis on PDMS surfaces in saturated n‐hexane and toluene vapor cannot be explained by a change in interfacial tensions. The observation supports the hypothesis that these vapors adsorb into the PDMS and form a lubricating layer. It is hoped that these findings help to solve fundamental problems and contribute to applications, such as anti‐icing, heat transfer, and water collection.

Fast removal of water drops from solid surfaces is important in many applications such as on solar panels in rain, in heat transfer, and for water collection. Recently, a reduction in lateral adhesion of water drops on poly(dimethylsiloxane) (PDMS) brush surfaces after exposure to various organic vapors was reported. It was attributed to the physisorption of vapor and swelling of the PDMS brushes. However, it was later pointed out that a change in the interfacial energies by vapor adsorption could also have caused low drop adhesion. To find out how strongly each effect contributes, contact angles of water drops on three hydrophobic surfaces in different vapors are measured. In water-soluble vapors, a substantial decrease is observed in contact angles. This decrease can indeed be explained by a vapor-induced change in the interfacial tensions. The very low contact angle hysteresis on PDMS surfaces in saturated n-hexane and toluene vapor cannot be explained by a change in interfacial tensions. The observation supports the hypothesis that these vapors adsorb into the PDMS and form a lubricating layer. It is hoped that these findings help to solve fundamental problems and contribute to applications, such as anti-icing, heat transfer, and water collection.

Introduction
The introduction of Young's equation (Figure 1a) started a quantitative treatment of contact angle phenomena. Young's equation predicts the equilibrium contact angle from interfacial surface energies: [1] γ LV cosθ = γ SV − γ SL . Here, γ LV , γ SV , and γ SL are the liquid-vapor, solid-vapor, and solid-liquid interfacial energies, respectively. It can be derived by a force balance: in equilibrium, the horizontal forces acting on the contact line need to balance. Despite the fact that Young's equation [2] is its surface tension changes due to adsorption and enrichment of the organic molecules. [5,15] That vapor adsorbs to the free solid surface and may decrease γ SV is also known. [2] The comment motivated us to study more in general the effect of organic vapors on contact angles and contact angle hysteresis of water drops on hydrophobic surfaces. We try to answer three questions. How do water contact angles change in different organic vapors (Figure 1b,c)? Can these changes be explained by changes of the interfacial energies? Can a change in γ LV , γ SV , and γ SL explain the low contact angle hysteresis observed on PDMS in n-hexane, cyclohexane, and toluene vapor?

Results and Discussion
To answer the questions, we measured the water contact angles on various hydrophobic surfaces in different vapor environments. As representatives, we choose three hydrophobic surfaces (Figure S1, Supporting Information); all surfaces were made as described in the literature. [12,16] 1) PDMS-coated silicon wafers-briefly, PDMS was grafted by immersing the silicon wafer into a solution which contained 40 mL toluene (with saturated water) and 1.4 mL dimethyldichlorosilane. 2) 1H,1H,2H,2H-perfluorodecyltrimethoxysilane (PFDTS)coated silicon wafers-PFDTS was coated using a chemical vapor deposition method, by putting silicon wafers into a vacuum desiccator (30 mbar) with 20 µL 1H,1H,2H,2Hperfluorodecyltrimethoxysilane added in the bottom for 3 h. 3) Teflon AF (amorphous fluoropolymer) on gold was prepared by dip coating; the sputter-coated gold glass slides were withdrawn from the solution (1 wt% Teflon AF1600 in perfluorotributylamine) at a constant speed of 10 mm min −1 . See the Supporting Information for preparation details. To suppress electrostatic effects, all substrates have a high dielectric permittivity or a conductive layer underneath. The root-mean-square surface roughness of all surfaces as measured by atomic force microscopy was less than 1 nm. As vapors, we used n-hexane, cyclohexane (CYC), toluene, dimethylsulfoxide (DMSO), ethanol, and tetrahydrofuran (THF). DMSO, ethanol, and THF are miscible with water. n-Hexane, CYC, and toluene are immiscible and only dissolve to a low concentration in water; the saturation concentrations are 9.5, 55, and 515 ppm, respectively. [17] To measure water contact angles in vapor environment, we used a goniometer (OCA 35, DataPhysics Instruments) equipped with a custom-made closed metal chamber ( Figure S2, Supporting Information), where two glass windows are inserted for visualization. The solvent was inserted at the bottom of the chamber. Inside the chamber, the surface was mounted in a way that it did not get into direct contact with the liquid solvent. The chamber was sealed for more than 10 min before conducting contact angle measurements to generate a stable vapor atmosphere. Videos of sessile drops in side view were recorded when changing water volume gradually (1 µL s −1 ) between 5 and 20 µL using a motorized Hamilton syringe. Advancing and receding contact angles were determined by fitting an ellipse model to the contour images.
When exposing water drops to vapor of cyclohexane or toluene, the advancing and receding contact angles did not change significantly (Figure 2a). In hexane, they increased slightly (≈4°). When using soluble vapors, the contact angles decreased. The higher the saturation vapor pressure, the more the contact angles decreased (Table 1). For example, the advancing contact angles on a PDMS surface in hexane vapor and air were 110° and 106°, respectively, while it was only a v θ = 87° and 73° in ethanol and THF vapor, respectively.
To answer the second and third questions, we measured the surface tension of water in different vapors using the pendent drop method (DSA100E, Krüss). [18] To generate a saturated vapor environment, the liquid was inserted at the bottom of a custom-made closed metal chamber ( Figure S2, Supporting Information). After waiting 10 min to reach saturation, a hanging water drop was introduced and imaged in side view. Real-time results were obtained every second by fitting the drop shape with the Laplace equation with the density of water using a built-in program. We report the water's surface tension 2 min after placing the drop into the chamber (Table 1). 2 min is the same time delay as used for contact angle measurements.
The surface tensions of water drops in ethanol, THF, and DMSO vapors vapor decreased to 37, 36, and 58 mN m −1 , respectively. These values are lower than the ones in n-hexane, cyclohexane, and toluene vapors, which were 67, 69, and 63 mN m −1 , respectively. The low effect in DMSO vapor is attributed to the low saturation vapor pressure of DMSO.
To answer the second question, we first analyze which effect a change of γ LV has, by keeping all other interfacial energies constant. Which contact angle is expected, if γ LV changes from the value in air to the one measured in vapor while keeping γ SV and γ SL constant? To answer this question,   Figure S3 (Supporting Information)) for n-hexane, cyclohexane, and toluene within the error of our measurements. This agreement is not surprising, considering that the contact angles were not much affected by the vapor. Substantial deviations were, however, observed for the soluble vapors ethanol and THF. Plotting the difference between the expected and measured contact angles (Figure 2c) shows that the measured contact angles tended to be lower than the expected contact angles. In particular, in THF vapor, the measured contact angles were 10°-23° lower than the calculated ones. Thus, a change from γ LV = 72 mN m −1 to the value measured in vapor LV v γ cannot explain the observed changes in mean contact angles for soluble vapors. We conclude that for soluble vapors the solid-vapor and/or the solidliquid interfacial energies change upon exposition to the vapor.
To be able to describe the changes in mean contact angles, we need to take into account a change of γ SV − γ SL . In particular, where v θ is the experimental mean contact angle in a specific vapor. On the left side of Equation (1), all parameters were determined experimentally. As shown in Figure 2d, Δ(γ SV − γ SL ) in all vapors is positive (except n-hexane). Specifically, in ethanol and THF, the value of Δ(γ SV − γ SL ) reaches more than 20 mN m −1 . It is highly unlikely that SV v SV air γ γ > , because a spontaneous adsorption of vapor to the solid surface should decrease the solid-vapor interfacial energy. [2,3] For this reason, we believe that this increase in Δ(γ SV − γ SL ) is mainly caused by an adsorption of dissolved vapor molecules to the solid-liquid interface, resulting in a decrease of γ SL .
To answer the third question, we investigate contact angle hysteresis in air  (Figure 3a and Table S1 (Supporting Information)). For the water-soluble vapors, Δθ v showed a trend to be higher than in air. For the nonsoluble vapors, the measured contact angle hysteresis Δθ v tended to be lower than in air. In particular, on the PDMS surfaces, much lower contact angle hysteresis was observed than in air.
In an attempt to explain the changing contact angle hysteresis by changing interfacial energies, we apply Young's equation locally. We assume that the interfacial energies close to the contact line need to be inserted into Young's equation and that the interfacial energies are different for the advancing and receding sides. The assumption of local validity of Young's equation is supported by the fact that just before the drop starts to slide, the forces acting on the contact line on the front and rear side are balanced. With this assumption, we estimate the "expected" advancing and receding contact angles in the vapor phase a, theo  Here, the interfacial energies in the first two equations refer to the ones close to the advancing contact line. We can write a similar equation for the receding side, assuming that to derive Equation (3), the interfacial energies close to the receding contact line need to be considered.
With Equations (2) and (3), we calculate the expected contact angle hysteresis in vapor (Figure 3b) and compared it to experimental results (Figure 3c). Experimental values for the contact angle hysteresis deviated substantially from the calculated values (Table S2 in the Supporting Information for details). For the water-insoluble vapors, the contact angle hysteresis tends to be lower than theoretical values. In particular, for the PDMS surfaces, the observed contact angle hysteresis was 71-96% lower than the theoretical ones. We attribute these low Δθ exp values to the lubricating effect of the vapor. Vapor is absorbed in the PDMS and lubricates the motion of the contact line. [20] We conclude, that the change of interfacial energies cannot explain the change of contact angle hysteresis. Physisorption into the PDMS plays an important role. This hypothesis is supported by the correlation of Δθ exp with the difference of Hansen solubility parameters [19,21] between the vapor and the PDMS coating (Δδ = δ V − δ PDMS ). Originally developed to decide if a substance is soluble in another, it has also been successfully applied to indicate the uptake of solvents by PDMS. [21] The lower the difference between Hansen parameters, the more vapor adsorbs in the PDMS brush and the lower the contact angle hysteresis become.

Conclusion
Water advancing and receding contact angles do not change significantly for vapors, which are immiscible with water. By contrast, for soluble vapors, we detected a substantial decrease of contact angles. This decrease can be explained by an adsorption of dissolved molecules to the solid-liquid interfaces and the resulting reduction of γ SV . Thus, in accordance with Khatir and Golovin, a change in the mean contact angle can be explained by vapor-induced changes in the interfacial energies. However, the observed decrease in contact angle hysteresis on PDMS cannot be explained by changes in γ LV , γ SL , and γ SV . We attribute them to an uptake of vapor by the PDMS and the formation of a lubricating layer. Our findings are directly relevant to situations where droplets occur such as water collection, heat transfer, and power generation applications. The insight that background vapor influences drop contact angles may open new avenues for drop manipulation.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.