Perspectives on superhydrophobic surface durability

Superhydrophobicity is a remarkable evolutionary adaptation manifested by several natural surfaces, such as lotus leaves,¹ pond skater legs,¹ and butterfly wings.² Extreme water repellency is achieved by combining micro- and nanostructures with conformal low-surface-energy materials. Bio-inspired artificial superhydrophobic materials offer exciting opportunities for self-cleaning,³ electrostatic energy harvesting,⁴ anti-dewing,^5,6 anti-icing,^7–9 and enhanced heat transfer.^10–16 Among these applications, condensation of water vapor has seen a particular focus because of its wide utilization in thermal management,¹⁷ air-conditioning,^18–20 and power generation.²¹ For example, to enhance condensation, researchers have recently proposed coalescence induced droplet jumping on rationally designed superhydrophobic surfaces,^22–27 where two or more condensate droplets coalesce on a low adhesion superhydrophobic surface and jump away from the surface. Traditional superhydrophobic surfaces rely on creating air-infused micro- and nanostructures to minimize the contact between the water and the substrate.^28,29 Although promising, the widespread use of superhydrophobic surfaces is restricted due to a lack of long-term durability, surface flooding at high humidity or droplet impact velociy,^31–33 droplet contact line pinning due to defects,^23,33,34 and loss of water repellency at elevated pressure and temperature.^35–39 

In this Perspective, the majority of our focus is on structured surfaces, which are intrinsically hydrophobic or are conformally coated to become superhydrophobic. Although traditional smooth hydrophobic or slippery liquid infused porous surfaces (SLIPS) can achieve similar water repellency to their superhydrophobic counterparts,^40–43 these liquid infused surfaces are only briefly discussed due to their large body of work. We explore the advancements made, challenges encountered, and future requirements for achieving durable superhydrophobicity. Using a top-down approach, we first discuss widely adopted coating chemistries for superhydrophobic surfaces. We then discuss current methods of fabrication and future guidelines for designing the underlying micro- and nanostructures on superhydrophobic surfaces. Throughout our discussion, we emphasize the significance of durability, scalability, as well as the challenges and potential for industrial implementation of superhydrophobic surfaces.

The combination of roughness and low surface chemistry directly control the degree of superhydrophobicity by successfully maintaining air gaps between micro- and/or nanostructures.³⁰ Classical artificial hydrophobic materials such as low surface energy polymers or self-assembled monolayers (SAMs) used to create superhydrophobic surfaces lack robustness.⁴⁴ Self-assembled monolayers (1–10 nm-thick) chemically degrade because covalent bonding with the substrate becomes chemically unstable when exposed to water.^45–47 In comparison, polymeric coatings (10–100 nm-thick) are chemically stable, but delaminate easily in humid environments due to the growth of underlying water blisters.^34,48 Strong coating adhesion is essential, and without a specific design goal for adhesion, introducing covalent chemical bonds between the coating and the substrate has become standard practice (with an estimated adhesion of ∼1 J/m²).^45,49,50 However, the design of covalent bonds has significantly limited the choice of polymer chemistries, making it an expensive and less versatile approach.

Recent studies on polymer coating delamination mechanisms have revealed that water blister formation is the primary cause of hydrophobic coating failure.^34,44,48,51 These findings highlight three crucial adhesion design guidelines: (1) the key factor is the underwater adhesion between the coating and the substrate (also referred to as the “wet adhesion”); (2) the apparent adhesion responsible for coating delamination is close to the thermodynamically reversible work of adhesion, which enables rational adhesion prediction before implementation using the classical Fowkes model;^52–54 and (3) the required adhesion to prevent delamination of a ∼100 nm-thick polymer film is only about 100–1000 mJ/m², which is weak and can be achieved even by non-covalent forces.^44,54 Taking these new insights into account, it becomes evident that covalent chemistries are not always necessary for achieving durable superhydrophobic surfaces.

Recent work employed a cell membrane-inspired coating that uses non-covalent adhesion to achieve stable jumping-droplet dropwise condensation for over a year, making industrial implication of dropwise condensation possible.⁵⁴ In nature, the phospholipid membranes of cells align with each other through non-polar hydrocarbon chains, leading to high underwater adhesion that prohibits biofluids to penetrate the interface and cause delamination.⁵⁵ By adopting a similar approach, one can employ a biomimetic double-layer coating that simply stacks one polymer coating on top of a non-polar SAM (Fig. 1). Even though neither coating material is robust when used alone, the polymer coating acts as a barrier to chemically protect the SAM from water exposure, and the SAM provides a lipid-like interface with high underwater adhesion, making the double-layer design very robust. This simple design principle allows for the theoretical prediction of a range of non-polar SAMs and polymers capable of providing satisfactory adhesion [Fig. 1(c)] and superhydrophobic durability during condensation.

Most polymers are electrically or thermally insulating, and some are thermally stable in a temperature range of −80 °C to 300 °C.⁵⁶ These characteristics allow polymers to sustain chemical, mechanical, thermal, electrical, or radiation exposure to some extent. Beyond these normal conditions, a new environment that pushes the capabilities of these materials to their limits can be considered as a “harsh” environment, which raises challenges for developing high-performance polymers, including the introduction of self-healing. Self-healing polymers refer to materials that can recover their structure and properties partially or fully after damage.⁵⁶ Self-healing materials are classified into two categories: extrinsic and intrinsic. In harsh environments, extrinsic self-healing that uses microcapsule-loaded healing agents in a polymer matrix has wide applicability and shows good self-healing efficiency.⁵⁷ However, this approach is irreversible and limited to a single healing operation. Extrinsic microvascular networks allow multiple self-healing events; however, their fabrication can be difficult, time consuming, and expensive. In comparison, intrinsic self-healing relies on macromolecular design of polymer structures.⁵⁸ High self-healing efficiencies and, theoretically, an infinite number of self-healing cycles can be achieved through dynamic covalent bonds or non-covalent interactions. However, to date, there have been limited reports where these techniques have been used in the development of polymers that can function and self-heal in harsh environments. In addition to the self-healing mechanism, the physical, thermal, and viscoelastic properties of the polymer must be carefully considered for each different environment.

Hydrophobic polymer materials have low thermal conductivity and elastic modulus (1–10 GPa), and poor mechanical stability in harsh conditions. To craft durable superhydrophobic coatings, past strategies have generally focused on optimizing the coating geometric and structural design. By doing so, these have demonstrated that coating robustness can be enhanced by surface structures (single- or multi-length scale), which act as a protective or sacrificial armor.^59–61 However, these hierarchical protective structures are substrate selective, not scalable, and the overall coating is thick (>10 μm) due to the height of the protective structures, not suitable for condenser applications. Although in recent years significant effort has been placed on designing thick and thermally conductive composite materials to overcome this challenge, scalable fabrication techniques for these coatings remain elusive.^62,63

Thus far, little progress has been made to enhance the superhydrophobic surface intrinsic properties to overcome the durability challenge. Recently, multiple studies have demonstrated atmosphere-mediated superhydrophobicity on rationally designed micro-/nanostructured surfaces, which enable jumping-droplet condensation without the need for synthesized coatings (Fig. 2).⁶⁴ The high-aspect-ratio CuO nanowires, together with the spontaneous adsorption of airborne volatile organic compounds (VOCs) that exist at a global scale, acquire condensation-tolerant superhydrophobicity when exposed to the atmosphere. Given the dynamic adsorption and an abundance of VOCs, exploiting VOCs as the coatings enables self-healing-based durable superhydrophobicity. Nevertheless, limitations to VOC coated surfaces exist. For instance, passive VOC adsorption is slow when compared to conventional functionalization methods^45,65 due to the inherently low concentrations of VOCs in the atmosphere. As seen in Fig. 2(b), it takes 3 weeks for the surface to reach saturation for the CuO nanowire surfaces exposed to the laboratory environment. Although plasma modification has been demonstrated to reduce the saturation timescale from weeks to days,⁶⁶ plasma cleaning adds additional vacuum-based processing steps to the overall manufacturing process. Along with these, the VOC adsorption process and rate also varies with surface structure.⁶⁶ 

The key to achieving high-quality superhydrophobicity is that a surface needs a large apparent advancing contact angle (>150°) and a low contact angle hysteresis (<10°). Many surfaces with excellent water repellency toward sessile and impinging large droplets (∼mm) undergo failure of superhydrophobicity during condensation due to the nucleation inside the roughness.⁶⁷ A large roughness factor, a higher intrinsic hydrophobicity, and a small length scale of the structures are conducive to maintaining good mobility of condensate droplets,²³ and a small solid fraction of the nanostructures minimizes the work of adhesion, thus promoting droplet jumping.^68,69 However, enhancing superhydrophobicity via the optimization of nanostructures has almost reached its limit, with minimum jumping droplet sizes approaching 1 μm.^6,68,69

Chemical etching and oxidation are two of the most scalable and adaptable (can be used for flat surfaces,⁷⁰ internal/external tube surface,⁷¹ and complicated finned geometries⁷²) techniques available for creating the roughness required for superhydrophobic surfaces. When fine-tuned properly, etching can produce hierarchical micro- and nanoscale structures, while oxidation techniques typically impart high aspect ratio nanometer-sized features (nano-grass,⁷³ knives,¹⁰ needles,⁷⁴ etc.) across the substrate. Oxide structures offer enhanced superhydrophobicity and higher condensation heat transfer coefficients by reducing the liquid–solid contact area, but this also leads to higher interfacial stresses and lower strength,^75,76 which makes them extremely fragile and susceptible to failure in common mechanical durability tests.⁷⁷ The inherent stresses due to droplet nucleation, movement of the contact line, and bulk drop removal can also lead to failure of structures [Figs. 3(a)–3(c)].⁷³ On the other hand, etched surfaces are generally more durable, as the micrometer scale and larger aspect ratio structures possess higher deformation strength,⁷⁸ and can also act as a shield to protect the nanostructures lying underneath in the valleys of adjacent micropillars, enabling the surface to retain its superhydrophobicity.⁷⁰ 

Currently, the scientific literature on durability of superhydrophobic surfaces faces two challenges: (1) a lack of standardized tests for measuring strength of micro-/nanostructures; and (2) a dearth of studies focusing on negative results and surface morphologies that fail to sustain prolonged superhydrophobicity. As a community, we should align our efforts to understand the failure mechanisms that are directly impacted by surface morphologies. Also, it should not be assumed that surfaces that show potential for simple stress conditions will survive harsher conditions prevalent in real applications. While the etching/oxidation process for aluminum and copper is well understood, focus needs to be placed on other substrates like stainless steel and titanium, which have notoriously shown resistance to surface modifications.

Similar to condensation, micro- and nanostructures have facilitated improved thermal performance for pool boiling^79–83 and flow boiling [Figs. 3(d)–3(f)].^84–88 However, a vast majority of these studies do not report the durability of these surfaces, thus rendering them impractical when it comes to industrial scale implementation. Structure integrity is particularly key for flow boiling where the significant liquid velocity can detach surface structures from the substrate. Two important factors to consider when analyzing structured surface durability for boiling are the structure size and oxidation effects. A factor to consider is the effect of growing oxide structures on the base metal for boiling enhancements. Due to the thermal expansion coefficient mismatch between the base substrate and the oxide structure that grows on it, thermal cycling can lead to different expansion rates. This, in turn, has the effect of eventual delamination of fabricated structures from the base metal. While the structure size effect has been hypothesized to be the primary factor driving performance degradation with time for nanostructured oxide-based surfaces, the effect of oxidation should also be considered and may be just as important.

In recent years, metal additive manufacturing (AM) has become a viable manufacturing technique due to its flexibility and capability of manufacturing complex shapes and designs.^91–94 As a result, scalable additive manufacturing techniques such as selective laser melting (SLM) has been utilized in manufacturing of highly efficient unconventional thermal management components having excellent thermal performance and power density [Figs. S1(a)–S1(d), supplementary material]. Components made with AM can be rendered superhydrophobic. Combining the advantages of surface functionalization with the design flexibility of AM components can have profound impact on a variety of industries.

To attain superhydrophobicity of AM surfaces, surfaces require well-defined rationally designed micro/nanostructures (see Fig. S1, supplementary material).^77,95 The microscale feature size that can be fabricated by selective laser melting (SLM) AM process is limited by the laser beam size which is about 60–80 μm and AM metal powder (i.e., AlSi10Mg uses powder of 20–63 μm distribution size).^72,95 This limits the possibility of direct AM of rationally designed nanostructured surfaces using SLM. However, the surface of the AM components can be micro-/nanostructured using post processing methodologies which will depend on the alloy material and the intrinsic surface sub-grain cellular microstructures [see Fig. S1(e), supplementary material]. Researchers have shown that through an understanding of grain formation mechanisms during AM combined with rational post processing methods, it is possible to develop previously unexplored unique micro-/nanostructures.⁹⁵ However, metal AM requires usage of nonconventional alloys, which adds new materials related challenges to surface micro- and nanostructuring. Recent developments of high-power laser have enabled additive manufacturing of variety of metallic materials, such as aluminum, copper, titanium, nickel, and stainless-steel alloys (i.e., AlSi10Mg, CuCrZr, Ti6Al4V, Inconel, SS 316L).^94–97 Hence, the techniques and the fundamental mechanisms governing the micro-/nanostructuring of various AM alloys need to be investigated for developing durable superhydrophobic surfaces.

In addition to the SLM AM process, researchers have been investigating other methods that are capable of direct manufacturing of smaller microscale features, i.e., hydrogel infusion,^98,99 jet electrolyte,¹⁰⁰ direct ink writing,¹⁰¹ and material jetting¹⁰² [see Fig. S1(g), supplementary material]. Although these methods provide more opportunities for fabrication of micro-/nano-structured AM superhydrophobic surfaces, most of these methods are at preliminary stage of development compared to SLM process, incapable of manufacturing large-scale thermal components, and need further investigation in the future.

Superhydrophobic surfaces have been extensively studied in water/steam applications, such as steam condensation, defrosting, and deicing.^103–105 However, few studies have investigated the behavior of other fluids such as synthetic (R-134a, R-245fa, etc.) or natural (CO₂, ammonia, etc.) refrigerants. One of the main reasons for this is that water has the highest surface tension among all liquids, except for mercury, making it easier for water to form hydrophobic droplet morphologies. In contrast, the surface tension of refrigerants is lower, typically on the order of ∼5 to 10 mN/m, making it challenging to create and characterize repellent surfaces using these fluids. Notably, measuring the contact angle of refrigerants presents a challenge, as these fluids tend to evaporate rapidly at atmospheric pressure. For instance, the contact angle of R-245fa was measured to be 20°,¹⁰⁶ on a copper surface using a high speed camera with a low-bond axisymmetric drop shape algorithm.¹⁰⁷ However, images were difficult to obtain due to the noise caused by the rapid evaporation of the refrigerant. Others overcame this challenge by building an environment chamber to measure the contact angle of R-134a on aluminum surfaces.⁸⁹ Although they showed droplet images, the quality of images is not as high as that of commercial goniometer used for water contact angle measurement.

Due to their low surface tensions, refrigerant can hardly form any non-wetting droplet morphology on metal surfaces. However, in certain applications, such as refrigerant boiling, superhydrophobic surfaces can still offer some benefits. Past work tested two different hydrophobic surfaces, namely, CuO and nanoFLUX, for refrigerant pool boiling.¹⁰⁶ The CuO nanocoated surface was hydrophilic (contact angle less than 5°, as measured with water) and later changed to hydrophobic (122 ± 24°), primarily due to the absorption of airborne VOCs.⁶⁶ While in the hydrophilic state, it suppressed refrigerant boiling heat transfer coefficient (HTC) by flooding potential nucleation sites. However, when the surface became hydrophobic, the HTC significantly increased, and the enhancement reached up to 50%. Meanwhile, the nanoFLUX surface was measured to have a water droplet contact angle of 161 ± 16° and resulted in a 200% HTC enhancement. The superhydrophobic surfaces enhance refrigerant boiling by providing more nucleation sites and inhibiting wettability. In contrast, for water boiling, superhydrophilic surfaces are more preferred due to the earlier onset of critical heat flux (CHF) when using superhydrophobic surfaces.^108,109

The industrial application of superhydrophobic surfaces faces significant challenges in terms of durability and scalability. Both structural integrity and hydrophobic chemistry are prone to damage, and achieving large-scale fabrication is a complex task. There is a pressing need for a wider range of fabrication techniques to create resilient structures with diverse properties. While aluminum surfaces are relatively easy to structure for traditional applications, copper is preferred for heat transfer purposes due to its superior thermal conductivity. Chemical etching or oxidation methods can generate random structures ranging from 1 to 10 μm, while laser etching or micromachining techniques can produce surfaces at scales ranging from 100 μm to sub-millimeter dimensions. However, there is limited research on fabricating well-defined copper microstructures within the range of >10 μm but <100 μm. One possible approach to fabricating large-scale porous structures is through dealloying. Dealloying is a common corrosion process in which an alloy is selectively dissolved, resulting in the formation of a nanoporous sponge composed predominantly of the more noble alloy constituents.¹¹⁰ For copper microstructuring, an alloy of copper can be fabricated with a less noble metal such as aluminum. Afterwards, by selectively removing aluminum from the Cu–Al alloy under controlled conditions, a wide range of structures can be fabricated. However, this technique can be expensive, and the resulting structures may lack mechanical robustness, potentially leading to material failure.¹¹⁰ 

Self-assembled monolayer functionalization is a popular approach for fabricating superhydrophobic surfaces. However, SAMs tend to fail in condensation applications. Similarly, recent studies have also shown that SAM performance declines during frost-defrost cycling, as it undergoes slow hydrolysis and accumulates particulate matter, leading to reduced superhydrophobicity.¹¹¹ Although SAM deposition is scalable, it lacks thermo-mechanical robustness. To overcome the challenges of SAMs or typical polymeric coatings, thinner coatings with a combination of scalable fabrication methods, higher thermal conductivity, good thermo-mechanical robustness, and long-term durability in humid environments are needed. One potential solution is the use of diamond-based coatings, such as diamond-like carbon (DLC), infused with low surface energy materials (fluorine). Fluorine infused DLC (F-DLC) not only demonstrates substrate versatility, enhanced dropwise condensation heat transfer, but also durability in moist environments for a period of more than 3 years.¹¹² Characterization of the compatibility of F-DLC in elevated temperature environments exceeding 300 °C and sustainability after 5000 mechanical abrasion cycles demonstrates resiliency.¹¹² However, the current design architecture of F-DLC is limited to smooth substrates or hydrophobic applications due to the similarity of the coating thickness with the length-scale of traditional microstructures. Future studies could focus on the development of superhydrophobic surfaces based on F-DLC.

To address the stability issues of superhydrophobic surfaces and their limited applicability to low surface tension liquids, researchers have proposed slippery liquid infused surfaces (SLIPS) or lubricant infused surfaces (LIS).⁴⁰ These lubricant infused surfaces can achieve low contact angle hysteresis (<3°) in combination with self-healing function due to the presence of the liquid layer utilizing capillary forces in the rough superhydrophobic structure. However, challenges remain, including the need to design immiscible and non-cloaking lubricants,^113,114 chemical instability of lubricant layer,¹¹⁵ toxicity of the lubricant used (typically fluorinated chemistries),¹¹⁵ and lubricant drainage with time.^116,117 Development of multifunctional SLIPS is needed to address these issues. Creating SLIPS with improved properties, possibly by using patterned SLIPS,^118–120 SLIPS with volumetric inclusion of the lubricant in the substrate,¹²¹ or combining superhydrophobic surfaces with SLIPS on well-designed structures, has the potential to overcome drawbacks and lead to enhanced performance and durability.

The development of superhydrophobic surfaces typically involves two main steps: structuring the substrate materials and applying hydrophobic chemistry. However, it is possible to fabricate superhydrophobic surfaces by micro-nanostructuring of the hydrophobic material itself, thus avoiding the need for processes, such as metal etching, oxidation, or dealloying. Past studies have shown patterning polymers as a route to develop superhydrophobicity.¹²² Adhesion of the polymer to the substrate and sustainable hydrophobicity pose challenges. Therefore, research should focus on developing techniques to etch the coating (to create structured coatings) pre-deposited on metal surfaces. In contrast, another possible route of water repellency is the generation of superhydrophobic surfaces without using any coating. Ion implantation is a promising technique for reducing surface energy and inducing hydrophobicity. Some studies have reported long-term condensation durability (∼5000 h) of hydrophobic surfaces achieved through ion implantation.^123,124 However, there is a lack of follow-up studies in this area, possibly due to the significant initial cost associated with the fabrication process. Nevertheless, recent advancements in technology have the potential to reduce capital cost, and surfaces created using ion implantation may overcome durability and scalability challenges. Unlike traditional coatings, ion implantation is a coating-less method and can be applied to large-scale surfaces (heat exchangers) after assembly, making it a viable option.

The absence of a standardized durability test(s) poses a significant challenge to the field of superhydrophobic surface durability. The current literature reveals a wide range of test protocols, making it difficult to compare findings and establish a consistent measure of durability. In condensation applications, it is crucial to conduct durability tests in controlled environments devoid of non-condensable gases (NCGs). This is because the presence of NCGs can have a substantial impact on condensation rates,¹¹² and different test environments may have varying concentrations of these gases, leading to inconsistent results that do not accurately reflect surface performance. To tackle this issue, it is essential to develop a method for predicting the lifespan and estimating the life cycle of superhydrophobic surfaces. Moreover, researchers should focus on highly accelerated life testing (HALT) of materials and establish protocols for durability prediction. It is important to note that laboratory-scale testing may not fully replicate real-world environmental conditions and performance. Therefore, greater emphasis should be placed on conducting tests in realistic scenarios, such as weathering tests that incorporate factors like condensation, exposure to UV light, and rain erosion. Such tests should aim to represent the specific application environments where superhydrophobic surfaces are used. Although weathering test chambers are commercially available, there is still a lack of data, prediction models, and standardized testing procedures that accurately capture material performance in specific environments. While there are some durability test results available for superhydrophobic surfaces in condensation environments, there is a notable dearth of studies and reported durability data for superhydrophobic surfaces in other phase-change applications. Standardizing durability tests and developing accurate prediction models will significantly contribute to the advancement and practical implementation of superhydrophobic surfaces, ensuring their reliability and performance in real-world applications.

The utilization of scalable deposition methods is crucial in reducing production costs and enhancing material efficiency. However, we should avoid material(s)/methods that involve solvents that emit VOCs and per- and polyfluoroalkyl substances (PFAS).¹²⁵ PFAS are a class of chemicals that do not naturally break down, and so they accumulate in water, soil, and in the human body. Studies have shown that high levels increase the risk of cancer and other adverse health effects.¹²⁶ The removal of PFASs from surface water, groundwater, soil, sediment, and biota is technically extremely difficult and very costly, if at all possible. Considering the effects of PFASs on human health and environment, in February 2023, European Chemical Agency (ECHA) banned around 10000 PFASs. By adopting environmentally friendly alternatives, we can mitigate the negative effects of these harmful substances and promote safer and more sustainable practices in the production of durable superhydrophobic surfaces.

In summary, superhydrophobic surfaces hold significant potential for various applications. However, further research is needed to develop scalable, durable, and environmentally friendly superhydrophobic surfaces suitable for industrial-scale use on a wide range of metals. Also, it is important for the community to prioritize the development of standardized durability testing protocols that focus on the end-use requirements of these superhydrophobic surfaces. These efforts will contribute to the widespread adoption of superhydrophobic surfaces and maximize their impact in societal applications.

See the supplementary material for details of the additive manufacturing (AM) and superhydrophobic surface development.

This work was funded by Naval Nuclear Laboratory under Agreement No. 147975, contracted to The University of Illinois. The authors gratefully acknowledge funding support from the Office of Naval Research (ONR) under Grant No. 014-21-1-2089. N.M. also gratefully acknowledges funding support from the International Institute for Carbon Neutral Energy Research (WPI-I2CNER), sponsored by the Japanese Ministry of Education, Culture, Sports, Science and Technology.

The authors have no conflicts to disclose.

Muhammad Jahidul Hoque: Writing – original draft (equal); Writing – review & editing (equal). Sujan Dewanjee: Writing – original draft (supporting). Nenad Miljkovic: Conceptualization (lead); Writing – review & editing (equal). Jingcheng Ma: Writing – original draft (supporting). Kazi Fazle Rabbi: Writing – original draft (supporting). Xiao Yan: Writing – original draft (supporting). Bakhshish Preet Singh: Writing – original draft (supporting). Nithin Vinod Upot: Writing – original draft (supporting). Wuchen Fu: Writing – original draft (supporting). Johannes Kohler: Writing – original draft (supporting). Tarandeep Singh Singh Thukral: Writing – original draft (supporting).

The data that support the findings of this study are available from the corresponding authors upon reasonable request.