Designing Lubricant-Impregnated Surfaces for Corrosion Protection

It is estimated that the total direct costs related to corrosion are $121 billion, or about 1.38% of the U.S. gross domestic product (GDP).¹  Protective coatings alone account for 88% of these costs, which include the raw materials to make the coatings as well as the costs associated with replacing damaged coatings on a yearly basis.¹  A vast majority of corrosion research has been devoted to developing coatings that can reduce corrosion.^1-2  These protective coatings act as barriers that reduce the contact between the corrosive media and the underlying metal surface, however, any mechanical defects such as cracks or holes formed over time within the coating can provide pathways for corrosion.

A lubricant-impregnated surface (LIS) is a hybrid coating whereby a textured scaffold is coated with a hydrophobic modifier and then impregnated with a liquid lubricant which is stabilized by capillary and intermolecular forces.^3-9  The hydrophobic modifier lowers the surface energy of the substrate and ensures that the lubricant is not displaced by water atop (see Figure 1), thereby making these surfaces remarkably slippery. The thickness of the lubricant is dictated by the size of the textured features as well as the lubricant viscosity, surface tension, and coating speed,^10-11  and a controlled process to make LIS typically results in an equilibrium lubricant film that strongly adheres to the baseline substrate. Various methods for texturing LIS substrates both at the micro- and nanoscale have been developed.^12-13 

Owing to their unique properties and composite liquid-solid features, LIS technology has found applications in minimizing ice adhesion,¹¹  drag reduction,^14-15  and mitigating scale formation.^16-17  LIS systems offer promise in corrosion protection because unlike solid coatings, holes and cracks do not form in lubricant layers. Numerous papers have studied combinations of lubricants and porous and textured surfaces under various corrosive environments.^18-22  However, it is unclear if the lubricant films used in these studies were specifically controlled for equilibrium thickness, which is only obtained if the LIS is prepared using a controlled impregnation process that carefully considers the lubricant’s properties and speed of application.^10-11  Lubricant films having thicknesses greater than equilibrium thin films may impart full surface coverage and provide temporary corrosion protection, but as thick films are not stabilized by capillary forces the lubricant can drain by gravity and other forces, consequently affecting corrosion protection.²³  In addition, many previous studies used a hydrophobic surface modifier such as Teflon^† that itself can contribute directly to the barrier resistance on the surface, thereby making it harder to isolate the true effect of the lubricant on corrosion protection. This warrants a more thorough investigation of LIS with systematic texture geometries and lubricant properties for corrosion protection, which is presented in this study.

Lubricant-impregnated surfaces were developed on textured silicon substrates. Silicon was chosen as the base material as it is relatively easier to pattern micropillars on silicon as compared to other materials such as steel. A silicon wafer (6 in [15.24 cm] diameter) was textured using photolithography, and the resulting micropillars had a cubical geometry with width a = 10 μm, height h = 10 μm, and pitch b = 10 μm (see Figure 1[b]).

Silicon is not typically susceptible to corrosion unless it is immersed in extremely strong bases or acids under high-temperature and pressure conditions, which are dangerous for a systematic corrosion study. Hence, in order to adapt the silicon surface to a more industrially-relevant material that can be tested for corrosion in 3.5 wt% sodium chloride, the silicon wafer was coated conformally with a 0.5 μm layer of iron using sputter-deposition. The wafer was then diced into 1 in by 1 in squares (2.54 cm by 2.54 cm) and the resulting test coupons mimicked the underlying texture on silicon. To confirm the susceptibility of the iron coating, both smooth and textured coupons were exposed to 0.1 M hydrochloric acid and observed under an optical microscope, as shown in Figure 2. The films rapidly deteriorated and bubbles evolved thereby confirming the suitability of these films for corrosion testing.

In order to design an LIS system for use in corrosion protection, (1) the lubricant should be immiscible with the corrosive test fluid, (2) the lubricant itself should be electrochemically inactive, and (3) the lubricant should preferentially wet the textured surface in the presence of the corrosive fluid. Two lubricants that have been previously shown in electrochemical studies to meet these criteria were selected in this research to prepare LIS: Krytox GPL 102^† (hereafter commercial lubricant) and silicone oil.²⁴  To reduce the surface energy of the test coupons and to make them suitable for impregnating with these lubricants, a special hydrophobic modifier namely fluorophosphonic acid (FPA) was applied. FPA forms a thin monolayer by binding with iron atoms,²⁵  and hydrophobicity was confirmed by measuring the contact angle of water on a smooth FPA-coated iron substrate: 98±3°. Following FPA deposition, the coupons were impregnated with the lubricant using a controlled dip-coating process described elsewhere,^10-11  thereby resulting in equilibrium lubricant films.

All potentiodynamic corrosion experiments were conducted using a Biologic VSP^† potentiostat in a three-electrode cell with 3.5 wt% sodium chloride as the test electrolyte and a standard Ag/AgCl reference electrode and an inert graphite counter electrode. The electrical connection was made on an area of the coupon that was not covered with FPA or lubricant. For all samples tested, a fixed geometric area of 0.5 cm² was immersed in the electrolyte. The micropillars provided additional surface roughness, which increased the geometric area of the substrate by a factor of 2.⁴  Hence, an actual electrochemically active area of 1 cm² was used for the corrosion current density calculations. After immersing in the electrolyte, the open-circuit potential (OCP) was allowed to stabilize for at least 1 h before testing. Potentiodynamic scans were conducted around the OCP with a 1 mV/s scan rate, and the corrosion current was determined from the Tafel fitting of the resulting plot.

The contact angle of the impregnating lubricant (o) on the solid (s) in the presence of the corrosive electrolyte (w), θ_os(w), characterizes how the lubricant preferentially adheres to the solid in the presence of the corrosive electrolyte.⁴  In the case of complete spreading of the lubricant on the surface, the lubricant is not only stably impregnated within the textures but also covers the tops of the texture as a thin equilibrium film, which results in θ_os(w) = 0°. This LIS state is called a “spreading LIS” (see Figure 3[a]). In contrast, in a “nonspreading LIS”, θ_os(w) > 0°. This results in the post tops being exposed (see Figure 3[c]). Nevertheless, for nonspreading LIS, stable impregnation within the textures is still possible and the specific thermodynamic criteria that govern the formation of these states based on the spreading coefficient have been described in previous literature.^4,23 

The well-defined nature of micropillars developed in this study allowed explicit identification of both spreading and nonspreading LIS systems. The contact angles of both the commercial lubricant and silicone oil were measured on a smooth iron coupon coated with FPA in the presence of 3.5 wt% NaCl (θ_os(w)). The commercial lubricant showed θ_os(w) = 0°, while silicone oil showed θ_os(w) = 30±2°. Therefore, with the commercial lubricant, a spreading LIS was obtained where the lubricant covered both the tops and inside the textures (see SEM image in Figure 3[b]), while silicone oil resulted in a nonspreading LIS. Discrete lubricant droplets were observed on top of the posts, as shown in Figure 3(d), indicating that the lubricant film was not stable on top of posts, thereby resulting in a nonspreading LIS.

Figure 4(a) shows the potentiodynamic polarization plots obtained using four test samples: (1) bare iron, (2) iron coated with the FPA hydrophobic modifier, (3) spreading LIS using the commercial lubricant, and (4) nonspreading LIS using silicone oil. Figure 4(b) compares the corrosion current densities as determined from the polarization plots. These currents were corrected for the ohmic resistance, which for both LIS cases represented less than 3% of the polarization resistance. It can be seen that both LIS systems—spreading and nonspreading—significantly reduce the corrosion current density relative to iron: from three orders of magnitude reduction with spreading LIS to one order of magnitude reduction with nonspreading LIS. It is also interesting to note that the corrosion current density of the iron coupon coated with the hydrophobic modifier FPA is found to be the same order of magnitude as that of iron. This confirms that any corrosion resistance characterization in this study for LIS is obtained only from the lubricant, and not from the hydrophobic modifier itself, in contrast with previous studies.

These results highlight the importance of determining the type of LIS toward ensuring effective corrosion protection. A spreading LIS with complete coverage of an equilibrium lubricant film provides superior corrosion protection as the lubricant always adheres to the surface both within and atop the textures. In contrast, even though a nonspreading LIS holds the lubricant stably within the textures, corrosion protection is reduced as the discrete drops atop the textures seen in Figure 3(d) are displaced by the corrosive electrolyte, thereby exposing the tops. A thicker lubricant film may temporarily provide corrosion protection but the lubricant layer on top of the posts is unstable and can drain over time. Therefore, it is important to systematically design LIS with full consideration of the lubricant’s spreading properties to ensure optimal corrosion protection.

To demonstrate the practical applicability of LIS, corrosion was studied on carbon steel, a commonly used industrial alloy, in a 3.5 wt% NaCl solution (brine). Microscale roughness was achieved by sandblasting a 1 in by 1.5 in (2.54 cm by 3.81 cm) carbon steel coupon, and the surface energy was lowered using FPA followed by dipcoating with the commercial lubricant. As shown in Figure 3(e), the LIS sample did not corrode even after 48 h of exposure. This shows that when systematically designed, LIS can provide protection even on an industrially relevant surface. Ideally, LIS is best suited for protection of static components (such as the interior of storage vessels) with periodic replenishment of the lubricant during maintenance cycles.

• In summary, corrosion protection on well-defined lubricant-impregnated surfaces has been systematically studied, finding that LIS offers superior corrosion protection when compared to bare iron in 3.5 wt% sodium chloride. It has been shown that this corrosion protection results exclusively from the protective lubricant layer rather than the hydrophobic barrier coating that is typically applied to ensure stability of LIS. Furthermore, it has been shown that the spreading characteristics of the lubricant is important toward optimal corrosion protection. When coated on carbon steel, LIS has been shown to provide lasting protection when exposed to brine.

The authors gratefully acknowledge financial support from NSERC Canada (PGS-D Award) and the MITEI-Equinor grant. The authors thank Thomas Allison and Jacky Chin for their assistance in surface fabrication and characterization.