FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES Dissertation zur Erlangung des Grades „Doktor der Naturwissenschaften“ im Promotionsfach Chemie Fachbereich Chemie, Pharmazie und Geowissenschaften der Johannes Gutenberg-Universität in Mainz Florian Geyer geboren in Worms Mainz, 2018 Die vorliegende Arbeit wurde am Max-Planck-Institut für Polymerforschung unter der Anleitung von und von Mai 2015 bis Dezember 2018 erstellt. Tag der mündlichen Prüfung: 04.04.2019 Dekan: 1. Berichterstatter: 2. Berichterstatter: 3. Berichterstatter: Dissertation an der Universität Mainz (D77) FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 5 TABLE OF CONTENTS ABSTRACT ..................................................................................................................... 7 ZUSAMMENFASSUNG ........................................................................................................ 9 1 INTRODUCTION ..................................................................................................... 13 MOTIVATION AND OUTLINE .................................................................................................. 13 HOW TO DESCRIBE WETTABILITY — THE YOUNG EQUATION ..................................................... 15 WETTING OF ROUGH SURFACES — THE WENZEL EQUATION ..................................................... 16 WETTING OF HETEROGENEOUS SURFACES — THE CASSIE-BAXTER EQUATION ............................ 17 SUPERHYDROPHOBIC SURFACES ........................................................................................... 17 SURFACE TENSION AND ENERGY ........................................................................................... 19 CONTACT ANGLE MEASUREMENTS ........................................................................................ 21 CONTACT ANGLE HYSTERESIS .............................................................................................. 23 ROLL-OFF ANGLE MEASUREMENTS ....................................................................................... 26 SUPER-LIQUID-REPELLENT SURFACES .................................................................................. 26 SILICONE NANOFILAMENTS .................................................................................................. 34 APPLICATIONS OF SUPERHYDROPHOBIC AND SUPER-LIQUID-REPELLENT SURFACES ................... 38 PARTICLES AT INTERFACES — LIQUID MARBLES ...................................................................... 43 2 HOW DROPS START SLIDING OVER SOLID SURFACES .................................................... 48 INTRODUCTION .................................................................................................................. 49 RESULTS AND DISCUSSION................................................................................................... 50 EXPERIMENTAL SECTION ...................................................................................................... 58 AUTHOR CONTRIBUTIONS .................................................................................................... 64 ACKNOWLEDGMENTS .......................................................................................................... 65 3 WHEN AND HOW SELF-CLEANING OF SUPERHYDROPHOBIC SURFACES WORKS ................... 66 INTRODUCTION .................................................................................................................. 67 RESULTS AND DISCUSSION................................................................................................... 68 EXPERIMENTAL SECTION ...................................................................................................... 80 AUTHOR CONTRIBUTIONS .................................................................................................... 87 ACKNOWLEDGMENTS .......................................................................................................... 87 4 ENHANCING CARBON DIOXIDE CAPTURE USING ROBUST SUPEROMNIPHOBIC MEMBRANES ... 88 INTRODUCTION .................................................................................................................. 89 RESULTS AND DISCUSSION................................................................................................... 91 CONCLUSION ..................................................................................................................... 98 EXPERIMENTAL SECTION ...................................................................................................... 98 AUTHOR CONTRIBUTIONS .................................................................................................. 104 ACKNOWLEDGMENTS ........................................................................................................ 105 6 TABLE OF CONTENTS 5 HOW TO COAT THE INSIDE OF NARROW AND LONG TUBES WITH A SUPER-LIQUID-REPELLENT LAYER ....................................................................................................................... 106 INTRODUCTION ................................................................................................................. 107 RESULTS AND DISCUSSION ................................................................................................. 108 CONCLUSION .................................................................................................................... 117 EXPERIMENTAL SECTION .................................................................................................... 117 AUTHOR CONTRIBUTIONS .................................................................................................. 125 ACKNOWLEDGMENTS ......................................................................................................... 126 6 POLYHEDRAL LIQUID MARBLES ............................................................................... 127 INTRODUCTION ................................................................................................................. 128 RESULTS AND DISCUSSION ................................................................................................. 128 CONCLUSION .................................................................................................................... 138 EXPERIMENTAL SECTION .................................................................................................... 138 AUTHOR CONTRIBUTIONS .................................................................................................. 143 ACKNOWLEDGMENTS ......................................................................................................... 143 7 IMPACT AND OUTLOOK .......................................................................................... 144 8 REFERENCES ....................................................................................................... 146 9 ACKNOWLEDGMENTS ............................................................................................ 164 10 APPENDIX ........................................................................................................... 165 SUPPORTING INFORMATION FOR CHAPTER 3 ......................................................................... 165 SUPPORTING INFORMATION FOR CHAPTER 6 ......................................................................... 169 11 CURRICULUM VITAE .............................................................................................. 173 12 LIST OF PUBLICATIONS AND CONTRIBUTIONS ............................................................. 175 PUBLICATIONS .................................................................................................................. 175 PATENT APPLICATIONS ...................................................................................................... 176 PRESENTATIONS (SELECTED) .............................................................................................. 176 FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 7 ABSTRACT Superhydrophobic, super-liquid-repellent, and superomniphobic surfaces have attracted much attention in industry and academia in the past two decades. Liquid drops on such surfaces take an almost spherical shape and easily roll off. Despite the substantial prog- ress, there is a lack of understanding of how drops move on such surfaces. While there are various approaches for the fabrication of these surfaces on open and flat substrates, the coating of complex geometries such as long tubes remains difficult. Crucial to the suc- cess of super-liquid-repellent surfaces is to devise novel and promising applications. This involves understanding the advantages and challenges of these surfaces for industrial or large-scale applications. To approach this task, this work combines fundamental and ap- plied aspects. On the one hand, the motion and lateral adhesion of drops on superhydro- phobic surfaces are investigated, for example, to gain an in-depth understanding of how self-cleaning works and when it fails. On the other hand, novel and improved surfaces are designed. The questions are motivated by environmental and social demands. Can carbon dioxide (CO2) be captured more efficiently using super-liquid-repellent membranes to re- duce CO2 emissions, e.g., in power plants? — To obtain a maximum packing of gas ex- change membranes, flat sheet membranes need to be replaced by porous super-liquid- repellent tubes. A method to fabricate meter-long super-liquid-repellent tubes is devel- oped. Long-term or biomedical gas exchange applications like blood oxygenation require anti-biofouling properties. Therefore, the antibacterial properties of the coated tubes are investigated. Finally, the design concepts are transferred to a different type of materials, the liquid marbles. Polyhedral liquid marbles are developed and characterized. The first part of this thesis answers the question of how drops start moving on solid hy- drophobic and superhydrophobic surfaces, in particular, which forces are involved dur- ing the onset and movement of the droplets. It is shown that the lateral adhesion force between a drop and a solid surface can be separated into a static and kinetic regime. This finding is unexpectedly similar to solid-solid friction. The force on the droplet increases until a specific threshold force is reached in the static regime. After overcoming the threshold force, the drop starts moving, and the friction force decreases to a constant value in the kinetic regime. This analogy is valid for various liquids of different surface tensions on smooth and rough surfaces. The second part investigates one of the most relevant applications and properties of su- perhydrophobic surfaces, originally inspired by the lotus leaf: self-cleaning. Contamina- tions such as dust are easily cleaned away by water drops rolling across these surfaces. Despite the vast interest in the self-cleaning properties, only little is known about the in- fluence of contaminant particle size, especially on the nanoscale and how the self-clean- ing process proceeds on the micrometer scale. It is shown that the pore size of the super- 8 ABSTRACT hydrophobic surface is crucial and needs to be smaller than the contaminant size to ob- tain contamination-resistant surfaces. The self-cleaning process is visualized on the mi- crometer scale using confocal microscopy, and the lateral adhesion forces while drops move over the contaminated surfaces are quantified, leading to a deeper understanding of the process. Outdoor environmental (>200 days) and industrial standardized contami- nation experiments show that the super-liquid-repellent surfaces used herein can resist contamination for prolonged periods. The third part focuses on the development of a novel application for superomniphobic surfaces that is the use of such a surface as a gas exchange membrane. A chemically, me- chanically, and thermally stable superomniphobic membrane for CO2 capture is fabri- cated. This membrane can repel highly basic, concentrated amine solutions commonly used for CO2 capture, e.g., in power plants. It provides a large liquid-gas interface and prevents wetting of the membrane, which usually reduces the gas exchange efficiency. This resulted in performance increasements of up to 40% relative to commercial mem- branes. In the fourth part, a strategy to coat the inside of meter-long and narrow tubes (down to 1 mm inner diameter) with a super-repellent-layer is developed. An oxidative Fenton so- lution is pumped through the tubes for activation, and subsequently, a super-liquid-repel- lent nanofilament coating is applied to the tubes. This method can be used to coat tubes made of various materials ranging from polymers, organic/inorganic hybrids, and ceram- ics to metals. The super-liquid-repellent coating is applied to medical catheter tubes, and their antibacterial properties are investigated. It is found that the nanofilaments show pronounced antibacterial properties and retain these irrespective of the wetting state. This makes the coated tubes a promising material for antibacterial catheters, and the coat- ing may be applied to porous tubes for gas exchange applications. The fifth part utilizes mobile superhydrophobic layers to encapsulate droplets, forming liquid marbles. Here, a new type of liquid marble is introduced, a so-called polyhedral liquid marble. Hydrophobic hexagonal plates adsorb to the liquid-air interface, stabilizing the droplet. The plates are separated by tiny air gaps, resembling a superhydrophobic micropillar layer. The plates can self-assemble into highly ordered hexagonally-arranged domains. Liquid marbles composed of only 4 to 6 plates with tetrahedral, pentahedral, and cubic shapes can be realized. Highly asymmetric polyhedral liquid marbles and let- ters are obtained due to the strong interfacial jamming facilitated by the rigid hexagonal plates. The polyhedral liquid marbles show various stimuli-responsive properties when exposed to water, ammonia, or tetrahydrofuran vapors. The gases are exchanged through the tiny gaps between the plates. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 9 ZUSAMMENFASSUNG Superhydrophobe, superflüssigkeitsabweisende bzw. superomniphobe Oberflächen er- zielten in den letzten zwei Jahrzehnten große Aufmerksamkeit in Industrie und Wissen- schaft. Auf diesen Oberflächen nehmen Flüssigkeitstropfen eine fast sphärische Form an und rollen sehr leicht ab. Trotz des großen Fortschritts in diesem Forschungsfeld ist die Bewegung von Tropfen auf solchen Oberflächen nicht gut verstanden. Während es viel- zählige Methoden zur Herstellung solcher Oberflächen auf offenen und flachen Substra- ten gibt, stellt sich die Beschichtung von komplexeren Geometrien, wie langen Schläu- chen, als schwierig heraus. Für den Erfolg dieser Oberflächen ist es essenziell, neue und vielversprechende Applikationen zu entwickeln. Dazu ist es wichtig, die Vorteile und Her- ausforderungen dieser Oberflächen für industrielle und großskalige Anwendungen zu verstehen. Um sich dieser Aufgabe anzunähern, kombiniert diese Arbeit fundamentale und angewandte Aspekte. Auf der einen Seite werden die Bewegung und laterale Adhä- sion von Tropfen auf superhydrophoben Oberflächen untersucht, um zum Beispiel ein tieferes Verständnis für die Frage, wie die Selbstreinigung dieser Oberflächen funktio- niert und was deren Grenzen sind, zu erlangen. Auf der anderen Seite werden neuartige und verbesserte Oberflächen designt. Die Fragenstellungen sind hierbei u. a. durch um- weltrelevante und soziale Erfordernisse motiviert. Wie zum Beispiel: Kann man CO2 mit Hilfe von superflüssigkeitsabweisenden Membranen effektiver abscheiden, um die CO2- Emissionen, beispielweise in Kraftwerken, zu senken? — Um eine maximale Packungs- dichte an Gasaustauschmembranen zu erhalten, müssen Flachmembranen durch poröse superflüssigkeitsabweisende Schläuche oder Rohre ersetzt werden. Hierzu wird eine Me- thode zur Herstellung von meterlangen superflüssigkeitsabweisenden Schläuchen entwi- ckelt. Langzeit- und biomedizinische Gasaustauschanwendungen wie beispielsweise Blutoxygenierung setzen Anti-Biofoulingeigenschaften voraus. Daher werden ebenfalls die antibakteriellen Eigenschaften der beschichteten Schläuche untersucht. Schließlich werden die Designkonzepte auf einen anderen Typ von Materialen übertragen, den Flüs- sigkeitsmurmeln. Hierbei werden sogenannte polyedrische Flüssigkeitsmurmeln entwi- ckelt und charakterisiert. Der erste Teil dieser Dissertation beantwortet die Frage, wie Tropfen sich auf hydropho- ben und superhydrophoben Oberflächen bewegen. Im Speziellen, was die beteiligten Kräfte während des Anfangs sowie während der Bewegung der Tropfen sind. Es wird ge- zeigt, dass die laterale Adhäsionskraft zwischen einem Tropfen und einer festen Oberflä- che in einen Haft- und Gleitbereich unterteilt werden kann. Diese Erkenntnis ist uner- warteterweise ähnlich zur Fest-Fest-Reibung. Im Haftbereich nimmt die Kraft auf den Tropfen bis zu einer gewissen kritischen Kraft zu. Nachdem diese kritische Kraft über- wunden ist, fängt der Tropfen an, sich zu bewegen und die Reibungskraft nimmt im Gleit- 10 ZUSAMMENFASSUNG bereich auf ein konstantes Niveau ab. Diese Analogie ist gültig für verschiedene Flüssig- keiten mit unterschiedlichen Oberflächenspannungen auf glatten sowie auf rauen Ober- flächen. Der zweite Teil untersucht eine der relevantesten Applikationen und Eigenschaften von superhydrophoben Oberflächen, welche ursprünglich durch das Lotusblatt inspiriert wurde: die Selbstreinigung. Kontaminationen wie Staub werden problemlos von Wasser- tropfen, die über die Oberflächen rollen, mitgenommen. Trotz des großen Interesses an den selbstreinigenden Eigenschaften, gibt es nur wenige Erkenntnisse über den Einfluss der Kontaminationspartikelgröße, speziell auf der Nanoskala und wie der Selbstreini- gungsprozess auf der Mikrometerskala abläuft. Es wird gezeigt, dass die Porengröße der superhydrophoben Oberfläche kleiner als die Kontaminantgröße sein muss, um eine ver- unreinigungsresistente Oberfläche herzustellen. Der Selbstreinigungsprozess wird auf der Mikrometerskala durch Verwendung von konfokaler Mikroskopie visualisiert und die lateralen Kräfte, während Tropfen sich über die kontaminierten Oberflächen bewegen, werden quantifiziert. Dies führt zu einem tieferen Verständnis des Selbstreinigungspro- zesses. Außenbewitterungs- (>200 Tage) sowie industrielle, standardisierte Kontaminati- onstests zeigen, dass superflüssigkeitsabweisende Oberflächen mit Poren im Nanometer- bereich Verunreinigungen im Allgemeinen widerstehen können. Der dritte Teil fokussiert sich auf die Entwicklung einer neuartigen Anwendung für su- peromniphobe Oberflächen: die Verwendung als Gasaustauschmembranen. Eine che- misch, mechanisch und thermisch stabile superomniphobe Membran für die CO2-Ab- scheidung wird hergestellt. Diese Membran kann stark basische und konzentrierte Amin- lösungen abweisen, welche zur CO2-Abscheidung, zum Beispiel in Kraftwerken, verwen- det werden. Die Membran verfügt über eine große Flüssigkeit-Gas-Austauschfläche und die abweisenden Eigenschaften verhindern, dass die Membran benetzt wird, was übli- cherweise zur Reduzierung der Gasaustauscheffizienz führt. Dies resultierte in Perfor- manceverbesserungen von bis zu 40% im Vergleich zu kommerziellen Membranen. Im vierten Teil wird eine Vorgehensweise zur Beschichtung des Inneren von meterlan- gen und schmalen Schläuchen (bis zu einem 1 mm Innendurchmesser) entwickelt. Eine oxidative Fentonlösung wird zur Aktivierung durch die Schläuche gepumpt und im An- schluss wird eine superflüssigkeitsabweisende Nanofilament-Beschichtung in den Schläuchen aufgebracht. Diese Methode kann zur Beschichtung von Schläuchen und Rohren aus unterschiedlichen Materialien, von Polymeren und organisch/anorganischen Hydriden bis hin zu Keramiken und Metallen, verwendet werden. Die superflüssigkeits- abweisende Beschichtung wird auf medizinischen Katheterschläuchen aufgebracht und deren antibakterielle Eigenschaften werden untersucht. Es zeigt sich, dass die Nanofila- mente ausgeprägte antibakterielle Eigenschaften besitzen und diese unabhängig vom Be- FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 11 netzungszustand erhalten bleiben. Dies macht die beschichteten Schläuche zu einem viel- versprechenden Material für antibakterielle Katheter und des Weiteren könnte die Be- schichtung auch auf porösen Schläuchen für Gasaustauschanwendungen aufgebracht werden. Der fünfte Teil benutzt mobile, superhydrophobe Schichten, um Tropfen einzuschlie- ßen, was zur Entstehung einer Flüssigkeitsmurmel führt. Hierbei wird ein neuer Typ von Flüssigkeitsmurmeln eingeführt, die sogenannten polyedrischen Flüssigkeitsmurmeln. Hydrophobe hexagonale Plättchen adsorbieren an der Flüssigkeit-Luft-Grenzfläche, um den Tropfen zu stabilisieren. Die Plättchen sind durch kleine Luftlücken voneinander ge- trennt, was einer superhydrophoben Mikrosäulenschicht ähnelt. Die Plättchen können sich zu hochgeordneten, hexagonalen Domänen selbstorganisieren. Es können Flüssig- keitsmurmeln bestehend aus lediglich vier bis sechs Plättchen mit tetraedrischen, penta- edrischen und kubischen Formen realisiert werden. Hochasymmetrische, polyedrische Flüssigkeitsmurmeln und -buchstaben werden aufgrund der starken Grenzflächenblo- ckierung durch die steifen hexagonalen Plättchen erhalten. Die polyedrischen Flüssig- keitsmurmeln zeigen verschiedene stimuli-responsive Eigenschaften, sobald man diese Wasser-, Ammoniak- oder Tetrahydrofurandämpfen aussetzt. Die Gase werden durch kleine Luftlücken zwischen den Plättchen ausgetauscht. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 13 1 INTRODUCTION MOTIVATION AND OUTLINE Superhydrophobic, super-liquid-repellent, and superomniphobic surfaces have attracted a lot of attention in industry and academia in the past decades. Since the 2000s, the inter- est in such surfaces has ever been increasing, currently peaking at almost 2000 papers published on superhydrophobic surfaces in 2018 (according to Web of Science, searching for “superhydrophobic”). These bio-inspired, self-cleaning surfaces are characterized by the fact that liquid drops show very low adhesion, take an almost spherical shape, and easily roll off the surface. Despite the large amount of research conducted, there are still many open questions and challenges to be addressed before these surfaces can be applied in real-world applications, which are vital for the success of these surfaces. The present thesis investigates both fundamental and applied aspects to contribute to the basic under- standing as well as prospective applications of super-liquid-repellent surfaces. Lateral adhesion force measurements are employed to reveal the onset of motion of drops on such surfaces (Chapter 2) and to characterize the forces needed to remove particle contamination on superhydrophobic surfaces. Self-cleaning is one of the most important properties and applications of superhydrophobic surfaces as the range of possible uses is large, e.g., for windows or as cover for solar cells. This work answers the question of what are the limits of self-cleaning and how to improve the surface design appropriately (Chapter 3). To fabricate super-liquid-repellent surfaces, this work generally employs a coating consisting of silicone nanofilaments. These nanofilaments have diameters be- tween 10 and 100 nm and lengths of up to several micrometers to introduce a surface roughness with re-entrant morphology needed for super-liquid-repellency. Superomni- phobic membranes based on nanofilament-coated fabrics are developed to answer the question if such membranes can be used to efficiently capture CO2, a highly relevant issue in times of record CO2 emissions and increasing global temperatures (Chapter 4). Here, it is particularly important that the membranes can super-repel highly concentrated amine solutions that are commonly used for gas scrubbing in industry. To obtain a maximal packing density of gas exchange membranes and thus a maximal exchange area per vol- ume, coated hollow fiber membranes are needed. Even though there is a large variety of methods to coat open and flat surfaces, ways to coat tubes with super-liquid-repellent lay- ers are limited. Here, a method to coat meter-long and narrow tubes (down to 1 mm in 14 INTRODUCTION diameter) is developed, and wide applicability to various materials is demonstrated (Chapter 5). Many long-term or biomedical applications like blood oxygenation demand anti-biofouling properties. Consequently, the anti-biofouling properties of the coated tubes are investigated for potential use in medical devices. This is important to answer another debated question in this field: Are superhydrophobic surfaces antibacterial? During self-cleaning, particles accumulate at the air-water interface. This can result in the formation of so-called liquid marbles. Liquid marbles are droplets encapsulated in a mobile superhydrophobic layer and are commonly prepared by rolling droplets on a (su- per)hydrophobic powder. The powder usually consists of spherical particles or particle aggregates. This work investigates if well-defined non-spherical particles can be utilized to fabricate such armored droplets and how they assemble on the surface of the drop (Chapter 6). Potential applications of these liquid marbles, including their response to external stimuli like different gases and vapors, are explored. This thesis1 gives a brief introduction to the basics of wetting and characterization of sur- faces in Chapters 1.2–1.9. Afterward, a brief overview of recent advances in the fabrica- tion of super-liquid-repellent surfaces (Chapter 1.10), including the synthesis of silicone nanofilaments (Chapter 1.11), is presented. Prospective applications of super-liquid-re- pellent surfaces, including self-cleaning, oil/water separation, anti-biofouling, and gas exchange, are discussed in Chapter 1.12. Particles at interfaces are described in Chapter 1.13, which is relevant for self-cleaning and the formation of liquid marbles. Chapters 2–6 present original contributions to the fundamentals, fabrication, and appli- cation of super-liquid-repellent surfaces and liquid marbles as well as the basic under- standing of droplet movement on such surfaces. Each chapter contains an Introduction, Results and Discussion, and an Experimental Section. The authors’ contributions are sum- marized at the end of each chapter. Chapter 7 discusses the impact of the findings and gives a brief outlook. 1The present thesis is a cumulative publication-based thesis, and parts thereof are published or submitted for publication in peer-reviewed journals. A note is added to the corresponding chapters. The Supporting Infor- mation is available online for the published manuscripts. Important Supporting Figures can be found in the Appendix for the submitted manuscripts. The full Supporting Information is available from the Author. Sup- porting Figures are generally labeled with the letter “S.” FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 15 HOW TO DESCRIBE WETTABILITY — THE YOUNG EQUATION2 Controlling the surface wettability[1-5] is essential in many areas of our daily life and indus- try. For instance, wettability determines the adhesion of drops[6-9] to window panes, paint to cars, and ice aggregation on airplanes.[10] Wettability is tuned in printing, microfluid- ics,[11] heat transfer applications,[12-16] and recently in the fabrication of metal-organic- frameworks.[17-21] Many natural surfaces, such as the wings of insects or birds, have unique nonwetting properties that are essential to their function.[22-23] The degree of wetting of a drop on a surface is determined by the interfacial tensions between the phases in contact, namely, the liquid, solid, and gas (air) phase. The fundamental equation correlating the contact angle of a drop with its interfacial tensions on an ideally flat, homogeneous, solid surface was developed by Thomas Young in 1805 (Eq. 1).[24] cos𝜃𝜃 𝛾𝛾LG + 𝛾𝛾SL = 𝛾𝛾SG (1) Here, 𝛾𝛾LG, 𝛾𝛾SL, and 𝛾𝛾SG are the liquid-gas, solid-liquid, and solid-gas 𝛾𝛾LGinterfacial ten- sions, respectively. The angle 𝜃𝜃 is called Young’s or intrinsic contact angle and is meas- ured at the three-phase contact line where the liquid, the solid, and the gas phase meet. The Young equation describes the balance of forces at the three-phase contact line be- tween the interfacial tensions acting parallel to the solid surface (Figure 1a). It should be noted that interfacial tensions are a force per unit length. Figure 1. a) Schematic illustration of a sessile drop. The Young contact angle at the three-phase contact line depends on the liquid-gas (𝛾𝛾LG), the solid-liquid (𝛾𝛾SL ), and the solid-gas (𝛾𝛾SG ) interfacial tensions. b,c) Sketch of a water drop on a hydrophilic surface (b) and on a hydrophobic surface (c). A liquid assuming a contact angle below 90° on a substrate is considered to be (partially) wetting, and a liquid with a contact angle of 90° and greater is considered to be nonwetting 2Parts of the introduction are submitted and accepted for publication within a review article in Adv. Mater. 16 INTRODUCTION (Figure 1b,c). Similarly, surfaces showing a water contact angle below 90° are defined as hydrophilic, while surfaces that show a water contact angle of 90° and greater are defined as hydrophobic.[5, 25] Young’s equation cannot be used to describe the contact angle of a sessile drop on rough or chemically inhomogeneous surfaces because the force balance at the three-phase contact line is affected by surface roughness or inhomogeneities. More complex approaches taking roughness and heterogeneities into account are needed. WETTING OF ROUGH SURFACES — THE WENZEL EQUATION The roughness factor r was introduced by Wenzel (Figure 2a) to account for surface roughness (Eq. 2):[26] Actual surface 𝑟𝑟 = (2) Apparent surface area The roughness factor is the ratio of the actual surface area divided by the macroscopically apparent surface area and is thus larger than unity. The contact angle on a rough substrate is given by the Wenzel equation: cos 𝜃𝜃W = 𝑟𝑟 cos 𝜃𝜃 (3) Here, 𝜃𝜃W is the apparent contact angle a drop forms on a rough surface, i.e., on a length scale much larger than the surface roughness or protrusions, and 𝜃𝜃 is Young’s contact angle of the material. In the Wenzel state, the liquid penetrates the pores of the rough surface. The introduction of surface roughness amplifies the intrinsic wetting properties of the material. On a hydrophobic material, roughness leads to a larger contact angle, whereas on a hydrophilic surface, roughness leads to a lower contact angle or complete wetting of the surface.[5, 26] Figure 2. Different wetting states. a) Schematic illustration of a droplet in the Wenzel state.[26] b) Schematic illustration of a droplet in the Cassie state assuming a solid-liquid-air com- posite interface.[27] FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 17 WETTING OF HETEROGENEOUS SURFACES — THE CASSIE-BAXTER EQUATION The wetting of composite surfaces consisting of different materials can be described us- ing the so-called Cassie-Baxter equation (Eq. 4):[27] cos𝜃𝜃C = �𝜙𝜙i cos𝜃𝜃i (4) 𝑖𝑖 Here, 𝜙𝜙i denotes the fraction of the surface in contact with the liquid that consists of ma- terial i, and 𝜃𝜃i is Young’s contact angle of material i. The Cassie-Baxter equation thus de- scribes the apparent contact angle 𝜃𝜃C based on the microscopic material composition. The so-called Cassie state is a situation in which a liquid droplet rests on the asperities of a rough surface (Figure 2b), entrapping air underneath the droplet. This solid-liquid-gas composite interface can thermodynamically be more favorable than wetting the rough structure (Wenzel state). The Cassie state can be described using 𝑖𝑖 = 2. Here, 𝜃𝜃1 = 𝜃𝜃 de- scribes the intrinsic contact angle of the solid material and 𝜃𝜃2 = 𝜃𝜃Air is the contact angle of a drop solely in contact with air (𝜃𝜃Air = 180°). The Cassie-Baxter equation can then be simplified to Eq. 5: cos 𝜃𝜃C = 𝜙𝜙SL cos 𝜃𝜃 + 𝜙𝜙Air cos 180° =𝜙𝜙SL(1 + cos𝜃𝜃) − 1 (5) Here, 𝜙𝜙SL = 1 − 𝜙𝜙Air describes the solid-liquid contact area. Such a composite interface usually results in contact angles far above 120°. Even though the Young, Wenzel, and Cas- sie-Baxter models and equations are important for the basic understanding of contact an- gle and wetting phenomena, it should be noted that these equations are derived assuming thermodynamic equilibrium between the three phases. Thus, none of the equations can account for pinning events or contact angle hysteresis (see Chapter 1.8 for a detailed dis- cussion) as these considerably complicate the theoretical and mathematical descrip- tion.[3, 28-29] SUPERHYDROPHOBIC SURFACES Extremely hydrophobic surfaces on which water drops adopt the Cassie state, show very high contact angles (𝜃𝜃C > 150°), and easily roll off when the surface is tilted by a few de- grees (<10°) are so-called superhydrophobic surfaces (Figure 3).[28, 30-32] The first superhy- drophobic surfaces were already described more than 50 years ago.[33] From the late 1970s 18 INTRODUCTION to the late 1990s, the biologist Barthlott and his co-workers shaped the term superhydro- phobicity based on investigations of the microscopic structure and wetting properties of numerous plants.[34] These findings initiated a substantial increase in academic and in- dustrial research on artificial superhydrophobic surfaces.[30] In nature, examples of su- perhydrophobic surfaces are lotus leaves (Nelumbo nucifera) or body parts of insects such as the legs of water striders, among others (Figure 3a).[34-35] Figure 3. Natural and artificial super- hydrophobic surfaces. a) Photograph of a lotus leaf. b,c) Micro- and nanostructure of the lotus leaf ob- served by scanning electron micros- copy (SEM). Reproduced from ref.[36] Copyright 2011, Beilstein-Institut, li- censed under CC BY 2.0. d) Model su- perhydrophobic surface consisting of SU-8 micropillar arrays. Reproduced from ref.[37] with permission from The Royal Society of Chemistry. e) Artificial lotus-like superhydrophobic surface consisting of raspberry-like nanoparti- cles. Reproduced from ref.[38] with per- mission from The Royal Society of Chemistry. Therefore, superhydrophobicity is often referred to as the lotus effect.[39] Besides the su- per-repellency, these surfaces are known for their exceptional self-cleaning properties.[31] Common to these surfaces is that they combine two unique features; a high surface rough- ness on the nano- or micrometer scale and low surface energy, i.e., an intrinsically hy- drophobic surface (intrinsic contact angle must close to or above 90°). The reason for the high water contact angles is that the solid-liquid contact area is very low[3], often even be- low 10% of the total composite interface. Due to the low solid-liquid contact area, water drops generally show low adhesion toward these surfaces and can easily roll off. Different FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 19 methods have been developed to prepare superhydrophobic surfaces, including the dep- osition of nanoparticulate coatings by spray[40] or dip coating,[38] candle soot deposition,[41] lithography,[42-43] chemical etching,[44] plasma etching,[45] and electrospinning.[46] SURFACE TENSION AND ENERGY The interfacial tension 𝛾𝛾 can be interpreted as the work d𝑊𝑊 required to generate a new surface area d𝐴𝐴 (Eq. 6):[5] d𝑊𝑊 𝛾𝛾 = (6) d𝐴𝐴 It takes units of energy per area (J m-2) or force per unit length (N m-1). It should be noted that the term “surface energy” is preferred for solids, whereas “interfacial tension” or “surface tension” are commonly used for liquids. Different interactions like van-der- Waals forces, electrostatic interactions, or hydrogen bonds lead to strong attraction be- tween molecules in the bulk. At the surface, parts of these interactions are absent, as the molecules at the surface are only partially surrounded by other molecules. Therefore, the transport of molecules from the bulk to the surface to increase the surface area requires energy to break intermolecular bonds. This energy is the surface energy or tension. Liq- uids’ surface tensions generally range between 20 mN m-1 (= 20 × 10-3 N-1) and 80 mN m-1.[5] For instance, water has a surface tension of 72.5 mN m-1 at 20 °C. However, some fluori- nated liquids have surface tensions as low as 12 mN m-1.[47-48] Surface tensions of liquids can be easily measured using the sessile or pendant drop method or with tensiometers utilizing a Du Noüy-ring[49] or a Wilhelmy-plate.[5, 50] However, measuring the surface en- ergy of a solid surface is less straightforward. The work required to deform a solid cannot easily be separated from the work that is required to create new surface area. Therefore, surface energies are mostly measured indirectly. There are several models for determin- ing the surface energy. Most of these models rely on the measurement of contact angles for liquids with different surface tensions. One of the first approaches is known as the Zisman plot (Figure 4).[51] Here, the cosines of the contact angles of various liquids are plotted against the respective surface tension of the liquid. It is important that the used liquids have largely different surface tensions to obtain a meaningful graph. The plot is extrapolated to cos 𝜃𝜃 = 1, which equals to an effec- tive contact angle of 0°, symbolizing complete surface wetting. This point represents the so-termed critical surface tension 𝛾𝛾crit and may be considered as the surface energy. After 20 INTRODUCTION all, this is not entirely accurate since the Zisman model does not account for the different types of surface interactions. The Fowkes[52] and Owens-Wendt-Rabel-Kaelble (OWRK)[53-55] methods introduce a distinction between dispersive and non-dispersive interactions and are therefore more suitable for polar surfaces. However, all of these methods are only applicable to materials with flat surfaces. They also assume that the drop is in thermody- namic equilibrium. Figure 4. Sketch of a typical Zisman plot. The cosines of the contact angles of differ- ent liquids on a surface are plotted against the surface tension of these liquids. Ex- trapolation of a linear fit to cos𝜃𝜃 = 1 yields the so-called critical surface tension 𝛾𝛾crit. Solid surface energies can be as low as 6 mN m-1 and as high as several thousand mi- cronewtons per meter.[56-58] Glass, metals, and metal oxides have high surface energies due to their strong (polar) intermolecular bonds and can form strong interactions with polar liquids, such as water. Conversely, low surface energy solids are nonpolar. There- fore, the intermolecular interactions of water are stronger than those between water mol- ecules and nonpolar surfaces. Thus, water droplets assume high contact angles on such surfaces to minimize the contact area with the solid surface. Highly fluorinated surfaces have particularly low surface energies. An ideal surface fully covered with -CF3 groups exhibits the lowest possible surface energy (6–7 mN m-1).[58] Exchanging one fluorine atom for hydrogen (-CF2H) already raises the surface energy to 15 mN m-1. A further exchange of fluorine for hydrogen yields even higher surface energies (Table 1). It was also found that surfaces covered with terminal groups (-CF3, -CH3) have lower surface energies com- pared to their nonterminal counterparts (-CF2-CF2-, -CH2-CH2-).[56] FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 21 Table 1. Surface energies for various surface compositions and surfaces at 20°C.[56-58] Surface composition/surface Surface energy [10-3 N m-1] -CF3 6–7 -CF2H 15 -CF2-CF2- 18 -CF2-CH2- 25 -CH3 20–24 -CH2-CH2- 31 -CH- (phenyl ring edge) 35 -CCl2-CH2- 40 MgO (100) 1200 Silicon (111) 1240 Thus, to obtain low surface energies and high hydrophobicity, a surface should have a high content of fluorocarbon compounds or at least large nonpolar hydrocarbon frag- ments. CONTACT ANGLE MEASUREMENTS Hydrophobic and superhydrophobic surfaces are typically characterized by conducting contact angle measurements with a goniometer. A liquid drop is deposited on the surface and observed using a camera from the side. The drop profiles of static or moving drops are then recorded.[59] The contact angle is extracted from a fit of the drop profile. The fit- ting method should be chosen with care because it can affect the extracted contact angle. Figure 5 shows the fitting of a droplet using four common methods: ellipse, circle, tan- gent, and Young-Laplace fitting. These methods yield vastly different results for the same recording (ranging from roughly 150° to 180°).[60] The circle and Young-Laplace fits assume that the contact angles at the left and right sides of the drop are identical. This is rarely true for rough surfaces because the three-phase contact line is likely to be pinned at pro- trusions. The elliptical and tangent fits permit different angles at the left- and right-hand sides of the drop profile. The tangent fit tries to match the slope of the meniscus close to the three-phase contact line. Not only the fitting method but also the camera settings can influence the results. Particularly important parameters include the illumination, bright- ness, contrast, and gamma.[32] 22 INTRODUCTION Figure 5. Side view of a sessile drop on a super- hydrophobic surface. The lines show fits of the drop profile using different methods. a) Ellipse fitting. b) Circle fitting. c) Tangent fitting. d) Young-Laplace fitting. The baseline is shown in red. The turquoise lines show the fitting result. Reproduced from ref.[60] with permission from The Royal Society of Chemistry. e) SEM image of a superhydrophobic SU-8 micropillar array. f) Opti- cal photograph of a water droplet on a superhy- drophobic micropillar array. g) Laser scanning confocal microscopy image of a water droplet on the same superhydrophobic micropillar array. Reprinted figure with permission from ref.[61] Copyright 2016 by American Physical Society. It is generally difficult to measure contact angles above 150° because the narrow gap be- tween the liquid and the solid greatly reduces the amount of light passing through the gap and reaching the camera, making it difficult to identify the real contact line. Conventional optical methods also typically cannot resolve details underneath droplets, e.g., whether the droplet is in the Cassie or Wenzel state.[32, 60] It has recently been shown that other techniques like laser scanning confocal microscopy can be used to measure contact an- gles on superhydrophobic surfaces more accurately (Figure 5e–g).[61-62] In contrast to op- tical methods, the air cushion underneath a droplet on a superhydrophobic micropillar array and the position of the contact line can be fully resolved (compare Figure 5f to Figure 5g). A drawback of this method is that it requires a transparent substrate. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 23 CONTACT ANGLE HYSTERESIS On real surfaces with finite surface roughness and heterogeneity, one must discriminate between the so-called static, advancing, and receding contact angles.[63-64] The static con- tact angle can be measured directly after depositing the droplet on the surface. However, due to pinning of the three-phase contact line, deformation of the surface, evaporation, or slow spreading of the liquid, the static contact angle may change over time. Contact angles can also be influenced by the way the drop is deposited on the surface. The static contact angle is thus not a clearly defined state despite its importance for surface charac- terization. It lies between the advancing and receding contact angles, which can be meas- ured using the following two methods. Figure 6. Advancing and receding contact angles can be measured by adding and withdrawing liquid from a droplet. a) During the addition of liquid, the droplet is expanding. After reaching the advancing contact angle, the contact line starts to Move. b) During the removal of liquid, the droplet is shrinking, and the contact line starts to recede just after reaching the receding contact angle. Reproduced with permis- sion from ref.[65] Copyright 2017, John Wiley and Sons. Firstly, by slowly adding and removing liquid from a deposited droplet with a rate of around 0.5 µL s-1 or less (Figure 6). During the addition of liquid and expansion of the droplet, the contact angle increases until the advancing contact angle 𝜃𝜃Adv is reached and the contact line starts advancing. Similarly, during the removal of liquid and contraction 24 INTRODUCTION of the droplet, the contact angle decreases until the receding contact angle 𝜃𝜃Rec is formed. The difference between the advancing and receding contact angles is called contact angle hysteresis (CAH):[59, 63, 65] CAH = 𝜃𝜃Adv − 𝜃𝜃Rec (7) Another way to access the advancing and receding contact angles is to deposit a droplet on a surface and gradually increase the inclination of the surface (Figure 7). Just before the droplet starts to slide or roll down the inclined surface, the advancing and receding contact angles are formed at the lower and upper contact side of the droplet, respectively. However, the droplet’s adhesion is often too high, preventing the drop from rolling off. In this case, the advancing and receding angles cannot be measured using this method.[32, 66] Figure 7. a) Schematic illustration of a drop on a substrate tilted by the inclination angle 𝛼𝛼. Just before the droplet starts rolling/sliding, the advancing and receding contact angles are formed on the droplet’s lower and upper sides. b) Experimental comparison between the static contact angle and the ad- vancing/receding contact angles formed on an inclined substrate. Reprinted with permission from ref.[66] Copyright 2009, American Chemical Society. The contact angle hysteresis can be ≥100° on some surfaces; however, it can also be as small as a few degrees on specific substrates showing low drop adhesion, such as super- hydrophobic[67] and liquid-like surfaces.[68] Nonetheless, also on superhydrophobic sur- faces, the contact angle hysteresis can exceed 20° because the advancing contact angle should be close to 180°.[61] For these surfaces, the measurement of roll-off angles is more reliable because it does not require the measurement of the position of the three-phase contact line. Several factors and phenomena can cause contact angle hysteresis. One is surface roughness, whose effects can be explained by considering a simple case involving a droplet that is advancing over a surface with a microscopic protrusion or bump FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 25 (Figure 8; left to right from Position A). The droplet advances with an intrinsic contact angle of 90° until it reaches the protrusion. It then jumps to the middle of the bump, where it can assume the intrinsic contact angle of 90° again (Figure 8; Position B). To advance further, the droplet needs to overcome the bump. It microscopically retains its intrinsic contact angle of 90° and slowly moves over the bump causing macroscopic pinning and apparent contact angles exceeding 90°, which is the advancing contact angle (Figure 8; Position B to C). After overcoming the microscopic bump, it can spread further, restoring its intrinsic contact angle of 90°, as there are no further obstacles in this example. The same effect occurs for the receding contact line, but in the opposite direction, leading to contact angles below 90°.[5] Figure 8. Schematic illustration of a droplet advancing over a surface with a small bump (left side) and a more detailed view of the movement over the bump (right side). Adapted with permission from ref.[5] Copyright 2003, John Wiley and Sons. Real surfaces can be imagined as surfaces having lower or higher densities of these mi- croscopic bumps, resulting in different contact angle hystereses. Another cause of con- tact angle hysteresis is surface inhomogeneity (either structural or chemical). Like sur- face roughness, these inhomogeneities can cause pinning of the contact line. The advanc- ing contact line may be pinned in areas exhibiting a higher lyophobicity (liquid-repel- lency), whereas the receding contact line may be pinned in areas exhibiting a lower lyo- phobicity and thus a higher lyophilicity. Besides surface roughness and inhomogeneities, dissolved molecules can increase hysteresis as they often adsorb to the contact line and hinder the advancing and receding of the droplet. However, as mentioned before, high surface roughness must not necessarily cause a high contact angle hysteresis as the effect of roughness depends on the wetting state. For the superhydrophobic Cassie state, the roughness decreases the hysteresis, whereas, for rough surfaces in the Wenzel state, sur- face roughness highly increases contact angle hysteresis (the example in Figure 8 resem- bles the Wenzel state).[5] 26 INTRODUCTION ROLL-OFF ANGLE MEASUREMENTS In addition to contact angle measurements, roll-off angle experiments (sometimes called sliding angle measurements) are commonly conducted to determine the wettability of a substrate.[39, 69] A drop is deposited on the surface, and the surface is slowly tilted until the drop rolls off (Figure 7), which will occur if the drop’s adhesion to the substrate is low enough to be overcome by the gravitational force acting on it. Therefore, the drop volume affects the tilting angle. Volumes of 5 to 10 µL are typically used. In particular, superhy- drophobic surfaces,[67] slippery liquid-infused surfaces,[70-71] and liquid-like surfaces[68] show low roll-off angles. The roll-off angle is related to the contact angle hysteresis ac- cording to the following equation:[72-77] 𝐹𝐹RA = 𝑚𝑚𝑚𝑚 sin𝛼𝛼 = 𝑘𝑘𝑘𝑘𝛾𝛾(cos𝜃𝜃Rec − cos 𝜃𝜃Adv) (8) Here, 𝑚𝑚 is the mass of the drop, 𝑚𝑚 is the gravimetric acceleration, and 𝛼𝛼 is the inclination of the surface, i.e., the roll-off angle. The factor 𝑘𝑘 depends on the shape of the three-phase contact line and is usually taken to be 1.[77] The contact width 𝑘𝑘 of the drop is measured perpendicular to the movement of the drop. The contact angle hysteresis appears in the (cos 𝜃𝜃Rec − cos𝜃𝜃Adv) term. Consequently, a low contact angle hysteresis is associated with a low roll-off angle. Roll-off angle measurements have the advantage of being fast and more reproducible than contact angle measurements because they do not depend on the illumination, the choice of the baseline, and do not require fitting. SUPER-LIQUID-REPELLENT SURFACES Superhydrophobic surfaces can only repel water and other aqueous solutions with high surface tensions (≈70 mN m-1). Polar liquids with lower surface tensions, such as surfac- tant solutions, and especially non-polar liquids like oils and hydrocarbons wet these sur- faces. The intrinsic contact angle of low surface tension liquids like oils is usually below 90° on flat surfaces, even on perfluorinated ones.[51] Therefore, the oil penetrates the pores of a superhydrophobic surface and assumes the Wenzel state. Usually, this leads to super- spreading as a high surface roughness combined with intrinsic contact angles much be- low 90° yield negligible apparent contact angles (≈0°). Capillary forces pull the liquid into the pores of the nano or microrough surface and cause a high spreading velocity.[78] There- FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 27 fore, these surfaces are usually called superhydrophobic and superoleophilic. This spe- cific wetting scenario can be utilized for oil/water separation[78-80] applications. The first approaches toward oil-repellent surfaces have been made as early as 1997 by Tsujii et al.[81] Ten years later, in 2007, Tuteja et al.[82] experimentally showed and explained the mecha- nism of super oil-repellent surfaces, i.e., superoleophobic surfaces. The wetting of rough structures by oils can be prevented by the introduction of overhanging morphologies or “re-entrant” structures. These overhanging morphologies act as an energy barrier such that the liquid cannot easily penetrate the pores of the rough structure. Examples of such overhanging morphologies are microfibers and model hoodoo-shaped micropillars (Figure 9a–d). The liquid is trapped at the lower side of the microfibers or at the lower side of the top face of the hoodoo-shaped micropillars, respectively. Here, the liquid can assume its Young's contact angle 𝜃𝜃 without further penetrating toward the bottom of the substrate. The drop rests on the re-entrant structures entrapping air underneath and thus assuming the Cassie state. The oil drop forms a large contact angle and can easily roll off the surface.[82] The criterion for superoleophobicity is commonly defined as showing high contact angles (>150°) and low roll-off angles (<10°) toward oils having surface tensions around 20–30 mN m-1. A surface that shows high contact angles and low roll-off angles toward aqueous and non-aqueous liquids is usually called a superamphiphobic or super- omniphobic surface. Therefore, in most cases, superoleophobic surfaces are also super- amphiphobic (amphiphobic expresses the combination of hydrophobic and oleophobic) or superomniphobic (expressing that “all” or most common liquids are repelled). It should be noted that the criterion for superoleophobicity, superamphiphobicity, and su- peromniphobicity is not as clearly defined and agreed on as the one for superhydrophobic surfaces. For instance, these terms do not clearly define which liquids are being repelled. Kota et al. suggested that superomniphobic surfaces should be able to super-repel hexa- decane in order to justify the terminology.[83] Herein, surfaces that repel water and aque- ous solutions with surface tensions around 70 mN m-1 are termed superhydrophobic. Sur- faces that repel liquids with lower surface tensions, such as diiodomethane (50.8 mN m-1),[48] blood (or blood plasma; ≈50 mN m-1),[84-85] or iodobenzene (39.7 mN m-1)[48] and below, are called super-liquid-repellent. Finally, surfaces able to repel hexadecane and liquids with even lower surface tensions are called superamphiphobic or superomni- phobic. 28 INTRODUCTION Figure 9. Design of superamphiphobic surfaces. a) Scheme of the solid-liquid-air interface on a surface consisting of microfibers. b) Scheme of the solid-liquid-air interface on a surface consisting of hoodoo- shaped micropillars (or micro-hoodoos). c,d) SEM images of a superamphiphobic surface consisting of hoodoo-shaped micropillars with squared (c) and round top faces (d). Reprinted from ref.[82] with permis- sion from AAAS. The robustness of the Cassie state on surfaces like electrospun microfibers depends on the morphology of the structures (e.g., the diameter of the fibers) and the spacing be- tween the re-entrant structures (e.g., the average spacing between the fibers). Tuteja et al. introduced two design parameters for superamphiphobic or superomniphobic sur- faces.[82, 86] Firstly, the spacing ratio 𝐷𝐷∗, which is given by Eq. 9 for microfibers: 𝐷𝐷 + 𝑅𝑅 𝐷𝐷∗ = (9) 𝑅𝑅 Here, 2𝐷𝐷 is the average edge-to-edge spacing of the fibers, and 𝑅𝑅 is the radius of the fibers. The spacing ratio 𝐷𝐷∗ directly governs the solid fraction 𝜙𝜙SL of the composite interface and thus the apparent contact angle 𝜃𝜃C. The second design parameter is the robustness factor 𝐻𝐻∗, which is given by Eq. 10 for microfibers. ∗ 2(1 − cos 𝜃𝜃)𝑅𝑅𝑙𝑙𝐻𝐻 = C (10) 𝐷𝐷2 Here, 𝜃𝜃 is the intrinsic contact angle, and 𝑙𝑙C = �𝛾𝛾LG⁄𝜌𝜌𝜌𝜌 is the capillary length. The robust- ness factor 𝐻𝐻∗ is a measure for the stability of the Cassie state against applied pressure. It depends on the geometry like the spacing and the radius of the microfibers, the intrinsic contact angle of the material, and the liquid properties. Both 𝑅𝑅 and 𝐷𝐷 directly affect the spacing ratio and robustness factor. It should be noted that increasing 𝐷𝐷 leads to a higher 𝐷𝐷∗ and thus 𝜃𝜃C, but highly decreases the robustness factor 𝐻𝐻∗. Therefore, a metastable FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 29 Cassie state may not be achievable for all surfaces having re-entrant features. Generally, 𝐻𝐻∗ and 𝐷𝐷∗ should be ≫1, and the surface should have sparsely distributed highly over- hanging structures. Tuteja et al. found 𝐷𝐷∗ = 9 and 𝐻𝐻∗ = 46 for octane on their electrospun microfibers.[82] Butt et al. suggested design principles based on force balance considerations for pillars consisting of sintered spherical particles (Figure 10).[87] It was shown that the depinning impalement pressure 𝑃𝑃, i.e., depinning from the lower side of the top particles, is given by Eq. 11: 2𝜋𝜋𝛾𝛾 𝑅𝑅 𝜃𝜃Micro 𝑃𝑃 = LG sin2 � Adv � (11) 𝐴𝐴0 2 Here, 𝑅𝑅 is the radius of the spherical particles, 𝜃𝜃MicroAdv is the microscopic advancing con- tact angle on the particles, 𝐴𝐴0 is the total area per pillar, which is equal to 𝑎𝑎2 for the unit cell of a square lattice with 𝑎𝑎 being the lattice constant (distance between adjacent pillars in a row). Thus, the impalement pressure critically depends on the ratio of 𝑅𝑅⁄𝑎𝑎2, which is similar to the robustness factor 𝐻𝐻∗ by Tuteja et al.[82] Figure 10. Design parameters for superamphiphobic surfaces consisting of pillars of sintered particles with radius R and a pillar-to-pillar spacing of a. In the Cassie state, the contact line is pinned at the lower half of the particles at the height H. Without external pressure, the liquid-air interface is planar. Upon ap- plying a force, the liquid-air interface bends downward. Reproduced from ref.[87] with permission from The Royal Society of Chemistry. The apparent receding contact angle 𝜃𝜃Rec at the rim of a drop moving perpendicular to a row of pillars can be related to the pillar structure using Eq. 12: 𝜃𝜃 = 𝜋𝜋 �1 − 2𝑅𝑅Rec cos2�𝜃𝜃MicroRec ⁄2�� (12) 𝑎𝑎 Here, 𝜃𝜃MicroRec denotes the microscopic receding contact angle on the particles. To obtain a high receding contact angle 𝜃𝜃Rec, the ratio 𝑅𝑅⁄𝑎𝑎 should be as low as possible for a fixed value of 𝜃𝜃MicroRec . This again conflicts the impalement pressure 𝑃𝑃. Thus, the ratio between 30 INTRODUCTION the radius of the particles and distance of the pillars must be tuned carefully to obtain high receding contact angles and a high impalement pressure.[87] Compared to superhydrophobic surfaces, there is a limited number of methods to synthe- size or fabricate superamphiphobic and superomniphobic surfaces. For instance, search- ing for “superamphiphobic surface” in Web of Science yields ≈340 results as of Dec 2018, whereas searching for “superhydrophobic surface” yields ≈11200 results. Nonetheless, different methods[83, 88] have been developed to fabricate such surfaces, including various etching methods such as wet chemical etching,[89] electrochemical etching,[81, 90-91] plasma etching,[92] and reactive ion etching;[93] other methods include candle soot deposition as a template,[41] (flame) spray coating,[94-96] dip coating,[97-98] spin coating,[99] electrospin- ning,[82, 86, 100] and electropolymerization.[101] In the following, a few prominent examples are presented. One of the early reports showing exceptional superamphiphobicity is based on a candle soot template. Candle soot was collected from a paraffin candle by holding a glass slide above the flame (Figure 11a). The black deposit consisted of fractal-like chains of carbon nanoparticles with diameters of ≈40 nm (Figure 11b,c). The loose structure was superhy- drophobic and quite fragile. Water droplets easily rolled off the candle soot-coated sur- face but took soot particles along. To strengthen the network, a 20 nm thick silica shell (Figure 11d,e) was deposited on the brittle soot by chemical vapor deposition (CVD). Therefore, a candle soot-coated glass slide was placed in a desiccator containing vials with tetraethylorthosilicate (TEOS) and ammonia for 24 h. To yield a transparent coating, the silica-coated glass slide was calcinated at 600 °C in air. During heating, the carbon was burned out, but the silica network mostly kept its integrity and turned transparent (Figure 11f). The hydrophilic coating was now fluorinated by CVD of 1H,1H,2H,2H-per- fluorooctyltrichlorosilane. After the fluorination, the coating was superamphiphobic, showing high contact and low roll-off angles for water and various oils like hexadecane and tetradecane (≈5° roll-off angle for 6 µL tetradecane droplets; Figure 11g,h). This su- per-liquid-repellency can be explained by the fractal-like overhanging structure of the candle soot template, preventing the liquids from penetrating the rough structure (Figure 11i). Hexadecane droplets (5 µL) impacting at a velocity of 𝑣𝑣 = 1 m s-1 could easily bounce off the coating (Figure 11j) without transitioning into the Wenzel state. After the initial impact, the drop bounced twice, followed by some oscillations until the drop came to rest.[41] FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 31 Pan et al. coated stainless steel wire meshes (Figure 12a) by electrospinning a mixture of polydimethylsiloxane (PDMS) and fluorodecyl polyhedral oligomeric silsesquioxane (F-POSS), followed by a curing step to crosslink the PDMS. Figure 11. Candle soot-based superamphiphobic Coating. a) Candle soot is deposited on a glass slide using a paraffin candle. b,c) SEM images of the candle soot deposited on a glass slide. The coating consists of carbon nanoparticle (≈40 nm diameter) chains. d,e) SEM images of the candle soot after application of a silica shell. f) TEM image of the hollow silica shells after calcination at 600 °C. g,h) A 2 µL water (g) and 5 µL hexadecane (h) drop on a candle soot-based superamphiphobic glass slide showing contact angles of 165° ± 1° and 156° ± 1°, respectively. i) Schematic illustration of a drop on a candle soot-based super- amphiphobic surface. The particle chain-like morphology acts as a re-entrant structure. j) Hexadecane droplet bouncing on a candle soot-based superamphiphobic surface. Reprinted from ref.[41] with permis- sion from AAAS. The crosslinked PDMS and F-POSS mixture had a solid surface energy of 𝛾𝛾𝑆𝑆𝑆𝑆 ≈ 12 mN m-1. It is argued that the F-POSS molecules preferentially migrate to the surface, lowing its surface energy (Figure 12b). The coated mesh had overhanging structures at different length scales. The wires of the mesh itself act as a re-entrant structure and are combined 32 INTRODUCTION with the microstructure of the electrospun coating (Figure 12c). The superomniphobic meshes showed extreme repellency against liquids down to 19.8 mN m-1 with roll-off an- gles below 3° for all liquids (Figure 12d). Various liquids having a low surface tension, including dimethylformamide (DMF), toluene, acetic acid, hexadecane, hexylamine, and PDMS, showed high contact angles (Figure 12e), and jets of these liquids could easily bounce off the surface (Figure 12f). The coating could also shield the underlying mesh from corrosive liquids, such as concentrated bases and acids, e.g., hydrochloric acid and sodium hydroxide solutions.[100] Figure 12. Superomniphobic stainless steel mesh by electrospinning a mixture of polydimethylsiloxane (PDMS) and fluorodecyl polyhedral oligomeric silsesquioxane (F-POSS). a) SEM image of the coated hier- archically structured superomniphobic steel wire mesh 70. b) Elemental mapping of fluorine on the su- peromniphobic mesh. c) Higher magnification SEM image of the overhanging electrospun microstructure on the mesh. d) Predicted and measured roll-off angles of various liquids (5 µL) on the superomniphobic mesh. The photograph depicts an ethanol drop rolling on the mesh at an inclination of 2°. e) Photographs of various liquid droplets, including dimethylformamide (DMF), toluene, acetic acid, hexadecane, hexyl- amine, and PDMS, on the coated mesh. The liquids were dyed. f) Photographs of jets of the same liquids rebounding off the superomniphobic mesh. Reprinted with permission from ref.[100] Copyright 2013, Amer- ican Chemical Society. Very recently, a simple spray method was developed by Pan et al. to fabricate superom- niphobic surfaces on various substrates. Here, 1H,1H,2H,2H-perfluorohexyltri- chlorosilane (PFHTS) and n-butyl cyanoacrylate (n-BCA) were mixed in dichloropen- tafluoropropane. This mixture could be sprayed on various substrates, such as polymers, metals, glass, wood, and fabrics. As an example, the coating was sprayed on PDMS disks. The spray coating yielded a hierarchal structure with re-entrant morphologies FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 33 (Figure 13a,b). The surface energy of the coating was 𝛾𝛾 -1𝑆𝑆𝑆𝑆 ≈ 12 mN m due to the perfluor- inated groups of PFHTS (Figure 13c). The superomniphobic coating was able to super- repel various liquids (Figure 13d) such as tetrahydrofuran (THF), DMF, a hydrofluoric acid (HF) solution, and even n-hexane (𝛾𝛾𝐿𝐿𝑆𝑆 ≈ 17.9 mN m-1). Figure 13. Superomniphobic coating based on spray coating of a mixture of 1H,1H,2H,2H-perfluorohex- yltrichlorosilane (PFHTS) and n-butyl cyanoacrylate (n-BCA). a) SEM image of a superomniphobic spray- coated PDMS disk. The inset shows an n-hexane droplet resting on the superomniphobic surface. Scale bars: 50 µm, 5 mm (inset). b) Higher magnification SEM images of the coating. The inset shows a cross- section of the hierarchical overhanging structure of the coating. Scale bars 5 µm. c) Chemical structure of PFHTS (left) and n-BCA (right). d) Photographs of drops of various liquids, including water, tetrahydrofu- ran (THF), DMF, hydrofluoric acid (HF) solution, and n-hexane, on the superomniphobic coating. e) Ultra- violet-visible (UV-Vis) spectra showing the transparency of the superomniphobic coating depending on the coating time. The inset shows nine spray-coated PDMS disks on a mobile phone screen. The disks were coated for 1–9 min as indicated by the numbers on the screen. Scale bar: 1 cm (inset). f) An n-hexane drop bouncing on the superomniphobic coating. The time between the frames is 6 ms. Scale bar: 5 mm. Re- printed from ref.[96] by permission from Springer Nature: Nature, Copyright 2018. 34 INTRODUCTION Additionally, the coating showed high transparency depending on the spraying time (Figure 13e). A transparency of 95% of the visible light was achieved using a spray time of 4 min and 20% PFHTS. Nonetheless, the coating was still super-repellent to n-hexane. The superomniphobic coating could not only repel resting but also impacting n-hexane drops. The impacting drops showed low adhesion on the surface and could bounce five times before coming to rest (Figure 13f).[96] While there are various methods to coat flat and open substrates, coating the inside of tubes with a super-liquid-repellent layer re- mains difficult, especially for polymeric tubes. Wet chemical or electrochemical treat- ments can fabricate superhydrophobic metal pipes.[102-104] However, these methods are limited to the respective metal. A “near”-superhydrophobic tube was fabricated by coat- ing a thick PVC tube (14 mm inner diameter) using a commercial liquid-repellent coating (Neverwet).[105] To circumvent the problems associated with coating procedures, in some works, tubes were rolled from a flat superhydrophobic substrate.[106-108] Other works em- ployed a rough solid template pipe, and a polymer was crosslinked or cured around this template.[109-110] These methods require to etch away the inner template pipe or require a radial tube expansion to yield a superhydrophobic tube, which considerably reduces the choice of polymers. Another approach inserted a mesh into glass tubes followed by ap- plying a commercial superhydrophobic coating.[111] In Chapter 5, a coating method based on silicone nanofilaments (Chapter 1.11) is introduced to overcome such limitations. Us- ing this method, meter-long and narrow tubes of various materials can be coated with a super-liquid-repellent layer. SILICONE NANOFILAMENTS A versatile approach to transform various substrates into superhydrophobic, superhydro- philic, or superamphiphobic surfaces is the silicone nanofilaments (SNF) coating.[67, 112] To coat a surface with these nanofilaments, the substrate is exposed to trichloromethylsilane (TCMS) in the gas phase[112-113] or dissolved in a solvent.[114-116] The gas phase or solvent must contain a certain water content to hydrolyze the TCMS slowly. For the solvent phase reac- tion, various non-polar solvents like toluene or n-hexane can be used.[115-118] A necessity for the nanofilament growth is a hydrophilic substrate, preferentially containing hydroxy groups.[112, 115, 119] Therefore, surfaces are usually activated using cleaning and oxidizing agents like piranha, basic surfactant solutions, or oxygen plasma treatments. The hydro- lyzed or partially hydrolyzed TCMS molecules react with hydroxy groups on the surface FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 35 inducing a polysiloxane (silicone) polymerization (Figure 14a). Due to steric reasons, a maximum of two silanol groups can attach to the surface. During the polymerization, the polysiloxane chains start crosslinking, leading to the growth of the nanofilaments. Hy- drochloric acid (HCl) gas is formed as a byproduct during the hydrolysis.[119-120] Diverse substrates can be coated, including glass, silicon wafers, cotton and PET fabrics, polyole- fins, metals, and ceramics.[80, 112-113, 118, 121-122] Besides TCMS, also other organosilanes, such as vinyltrichlorosilane,[119-120] or a mixture of tetraethoxysilane/tetrachlorosilane with an organosilane (e.g., long-chain alkylsilanes or perfluoroalkylsilanes) can be used.[123-124] The nanofilaments possess high chemical and environmental stability.[125-126] The length and thickness of the nanofilaments can vary and depend on many parameters like the TCMS concentration, water concentration, temperature, solvent, activation, and the sub- strate (Figure 14b).[114-115, 127-129] Figure 14. Silicone nanofilaments. a) Reaction scheme for the formation of silicone nanofilaments. Tri- chloromethylsilane (TCMS) hydrolyzes due to trace amounts of water and reacts with hydroxy groups on the surface. A polysiloxane polymerizes on the surface and eventually starts crosslinking.[116, 118-120] Repro- duced with permission from ref.[118] Copyright 2018, John Wiley and Sons. b) Nanofilaments on glass slides coated at different water concentrations using TCMS in toluene. Reproduced with permission from ref.[115] Copyright 2011, John Wiley and Sons. The mechanism of the nanofilament growth is not fully understood due to the inherent complexity of the self-assembly. There are several suggested mechanisms proposing quite different approaches.[119-120, 130-131] An early approach by Rollings et al. proposed a diffusion-controlled mechanism for the gas-phase reaction. The hydrolyzed silanes react with hydroxy groups on the surface, act- ing as seeds for further nanofilament growth. The high density of hydroxy groups on a glass surface, for example, should limit the mobility of the silanes on the surface, leading 36 INTRODUCTION to small islands growing on the seeds. These small islands start extending into the gas phase where they are more exposed compared to the valleys in between. The diffusion of the silanes to the tips of these islands is more efficient, leading to preferential one-dimen- sional growth at the tip. Consequently, nanofilaments are eventually formed on the sur- face.[119-120] Figure 15. Proposed growth mechanisms for the silicone nanofilaments. a–c) Uniaxial elongation of a partly crosslinked polysiloxane film. d) SEM of hollow silicone nanofilaments. Reproduced with permis- sion from ref.[131] Copyright 2014, ICE Publishing. e–h) Droplet assisted growth and shaping (DAGS) mech- anism. Reproduced with permission from ref.[128] Copyright 2017, John Wiley and Sons. Korhonen et al. assume that first a flat polysiloxane film is formed on the surface of the substrate (Figure 15a–d). The gaseous byproducts of the hydrolysis and condensation re- action (HCl and H2O) preferentially migrate into defects inside of the film and form bub- FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 37 bles inside of it. The bubbles expand and elongate in one direction as the lowest cross- linking density is at the upper side of the cavity inside of the film. Thus, the formation of nanofilaments is caused by a “pressure-induced uniaxial elongation of a partially cross- linked polysiloxane film.” This mechanism eventually yields hollow silicone nanofila- ments (h-SNF).[131] Recently, Artus et al. proposed a so-called droplet assisted growth and shaping (DAGS) mechanism (Figure 15e–h). Nano- or micrometer-sized droplets are formed on the sur- face due to humidity in the gas phase or water dispersed in the solvent and chemical or topographic inhomogeneities on the substrate. It is argued that these droplets are stable since their activity is lowered by the absorption of various chemical species such as HCl, silanols, and siloxanes, which are formed during the hydrolysis of TCMS. These micro- scopic drops act as a spatial confinement for the polysiloxane polymerization. The poly- condensation at the liquid-solid interface inside of the droplet leads to one-dimensional growth and, finally, to the peculiar nanofilament structure.[128-129] The methyl groups of the polysiloxane polymer orient to the outside, i.e., the surface of the nanofilaments, to minimize the surface energy. The high surface roughness of the coating and the intrinsically hydrophobic properties of the nanofilaments result in the superhydrophobic properties.[112-113] To obtain superoleophobic properties, the surface energy of the nanofilaments must be further reduced. Therefore, the nanofilaments are first activated in an oxygen plasma to oxidize their surface, forming hydroxy and other functional groups. In this oxidized state, the coating is superhydrophilic (and superoleo- philic) due to the high surface roughness and hydrophilized nanofilaments. This leads to the spreading of liquid droplets (contact angle ≈ 0°), in agreement with the Wenzel equa- tion. After activation, the nanofilaments are modified with a perfluoro silane (usually 1H,1H,2H,2H-perfluorooctyltrichlorosilane [PFDTS] or 1H,1H,2H,2H-perfluoro-decyltri- chlorosilane [PFOTS]).[115, 132-133] After modification, the surface energy of the nanofila- ments is highly reduced, minimizing dispersive and non-dispersive interactions. The fluorinated nanofilaments usually exhibit super-liquid-repellent or superamphiphobic (superomniphobic) properties. Depending on the specific structure of the nanofilaments, i.e., the thickness and spacing between the nanofilaments, the repellency toward low sur- face tension liquids, such as oils, may vary.[115, 133] Nonetheless, superhydrophobic prop- erties are (almost) always achieved. The super-liquid-repellency can be explained by the cylindrical geometry of the nanofilaments acting as re-entrant structures (as described in Chapter 1.10) similar to the electrospun fibers by Tuteja et al.[82] 38 INTRODUCTION APPLICATIONS OF SUPERHYDROPHOBIC AND SUPER-LIQUID-REPEL- LENT SURFACES Since a large variety of methods to fabricate superhydrophobic and super-liquid-repellent surfaces has been established, it is important to utilize their unique properties and abili- ties in our daily life or the industry. So far, prospective applications of super-liquid-repel- lent surfaces include particle synthesis,[134-138] water-proof textiles,[79, 139] oil/water separa- tion,[20, 78-80, 140-142] anti-icing,[143-144] self-cleaning,[145-146] drag-reduction,[102, 147] anti-biofoul- ing,[118, 148] gas exchange,[84, 116] membrane distillation,[149-150] and chemical shielding.[96, 100] In the following, a brief overview containing selected examples is given. Exposition of surfaces to outdoor environments inevitably results in contamination by particulate matter. There are different types of particulate matter from natural (e.g., mi- croorganisms, pollen) and anthropogenic sources (soot from combustion or other indus- tries). The size distribution of particulate matter is broad. It can range between tens of nanometers to hundreds of micrometers.[151] Superhydrophobic and super-liquid-repel- lent surfaces are known for their self-cleaning behavior (Figure 16). Water drops can eas- ily clean contaminated superhydrophobic surfaces. In contrast, contamination particles on common hydrophobic surfaces usually stick to the surface and cannot be simply cleaned by rinsing with water droplets. On superhydrophobic surfaces, macroscopic con- tamination particles rest on top of the rough structure. Therefore, macroscopic contami- nation can be easily removed by rinsing with water drops.[31, 34, 145, 152] Even though self- cleaning is one of the reasons for the huge interest in superhydrophobicity, only a few articles are dedicated to self-cleaning performances. Figure 16. A drop on a hydrophobic surface moves across the contamination particles being unable to pick them up (left). On a superhydrophobic self-cleaning surface, the drop easily picks up the contamination and takes it away. (right). Reproduced from ref.[152] with permission from The Royal Society of Chemis- try. For instance, previous research investigated the contamination of superhydrophobic sur- faces with millimeter- to micrometer-sized contaminants such as hydrophobic powder,[145] micrometer-sized particulates,[153-155] and aggregated bioparticles.[156] As an example, Fürstner et al. studied the wetting and self-cleaning properties of artificial superhydro- phobic surfaces, including silicon wafers with regular micropillars, microrough metals, FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 39 and replicas of superhydrophobic plants.[145] The surfaces were contaminated by blowing a hydrophobic powder (a mixture of a luminescent material and a fluorinated agent) onto the surface using a fan.3 Subsequently, the surfaces were cleaned using artificial fog (8– 20 µm droplets, 1500 mL m-2) and artificial rain (water drops larger than 2 mm). The fog was less efficient for cleaning the surfaces compared to artificial rain. Especially on sam- ples with pronounced structures above 5 µm, the fog treatment yielded a liquid film on the surface having no cleaning effect. The other samples with smaller structures were efficiently cleaned by the fog. The artificial rain led to almost “perfect” cleaning. Here, higher impact pressures showed more efficient cleaning performances. In contrast, a smooth surface could not be entirely cleaned by impacting droplets. However, it is still unclear how surfaces are affected if they are contaminated with particles of smaller sizes, especially in the nanometer regime, how the self-cleaning process works in detail, and which forces are involved. Chapter 3 will give an answer to that. Super-liquid-repellent surfaces can be utilized for the fabrication of various types of mi- croparticles.[134-135, 137-138, 157] This may contribute to a greener synthesis of such particles as no toxic solvent is needed. A monomer drop, e.g., consisting of methacrylates and con- taining a photoinitiator, is deposited on a super-liquid-repellent surface (Figure 17). Figure 17. Solvent-free particle synthesis on a superamphiphobic substrate. a) A monomer drop (here: a mixture of bisphenol A glycerolate dimethacrylate (bis-GMA), triethylene glycol di- methacrylate (TEGDMA), and the photoinitiator phenylbis(2,4,6-trimethylbenzoyl)phosphine oxide) is deposited on the surface and exposed to UV irradiation. b,c) Photographs of resultant polymerized microparticles prepared with a mixture of bis-GMA (15 wt%), TEGDMA (84 wt%), and a photoinitiator (1 wt%). d) SEM image of a microsphere prepared with a mixture of TEGDMA (99 wt%) and a photoinitiator (1 wt%). The monomer droplet was deposited using an inkjet printer. Reproduced with permission from ref.[134] Copyright 2013, John Wiley and Sons. Due to the super-liquid-repellency, the drop shows an almost spherical shape having a high contact angle >150° and a low roll-off angle. Consequently, the drop can easily roll 3Powders are highly aggregated. Thus, it is likely that the average aggregate size was above 10 µm. The authors mention that no fragments or particles below 1 µm were found. 40 INTRODUCTION on the super-liquid-repellent substrate. The drop is exposed to UV light to initiate cross- linking. During the UV exposure, the drop should be kept in motion to avoid deformation due to adhesion at the contact line or due to gravity. This can be realized by employing a curved super-liquid-repellent watch glass on a 2D orbital shaker (Figure 17a). After polymerization, spherical and solid microspheres are obtained (Figure 17b–d). Alterna- tively, polymer powders are melted on a super-liquid-repellent surface. After the transi- tion from the solid to the liquid state, the melted polymer assumes a spherical shape and can be cooled down, obtaining a polymeric microsphere.[134] Another approach utilizes aqueous solutions of titania nanoparticles or other metal ox- ides to fabricate inorganic mesoporous supraparticles on a super-liquid-repellent sub- strate. The particles are formed by slowly evaporating the aqueous solution. During evap- oration, the drop keeps its spherical shape until the water is completely evaporated, and a mesoporous inorganic particle is left.[157] Oil spills on the oceans by offshore oil production and transportation, for example, are frequent problems and can cause severe damage to humans, nature, and animals.[80] One of the most highly covered applications is oil/water separation using superhydrophobic membranes and sponges.[20, 78-80, 140-142, 158] This is achieved by the fact that most superhydro- phobic surfaces are superoleophilic, i.e., the surface repels water, but oils easily spread on it. Figure 18. Oil/water separation using a superhydrophobic/superoleophilic electrospun membrane. An oil/water mixture is poured on top of the membrane. The oil (red) passes through the membrane, but the water (blue) is repelled and stays on top of the membrane. Within 30 s, the oil/water mixture is separated. Reproduced from ref.[140] with permission from The Royal Society of Chemistry. The principle is simple; a porous superhydrophobic/superoleophilic membrane is placed between two columns (Figure 18). An oil/water mixture is poured on top of the membrane. The oil spreads on the superoleophilic membrane, and gravity pushes the oil through the FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 41 membrane. Nonetheless, the water phase stays on top of the membrane as it is water- repellent. Thus, oil can be efficiently separated from the water.[140] One of the drawbacks is that a small amount of oil usually remains in the water. Superhydrophobic/superoleo- philic sponges were used to pick up oil spills from water surfaces. Due to the superhydro- phobic properties, the sponge can float on water without taking up water. However, as soon as the sponge contacts oil, the oil spreads on the sponge’s surface and gets sucked inside. This proved to be an efficient way to remove oil spills from a water surface.[158] Another promising feature of superhydrophobic surfaces is that they may have antibac- terial properties.[118, 148, 159-164] Formation of biofilms,[165-166] i.e., the irreversible attachment and colonization of bacteria on surfaces like medical devices, can lead to infections and eventually cause the death of patients. Also, due to the increasing resistance of bacteria against common antibiotics,[167] it is crucial to control and reduce the bacterial adhesion on such devices. Bacterial solutions on superhydrophobic surfaces show high contact an- gles and easily roll off because of the low solid-liquid contact area in the Cassie state. Due to the few contact points with the superhydrophobic surface, bacteria can only attach to a few sites on the surface. This can lead to highly reduced bacterial adhesion and preven- tion of biofilm formation on superhydrophobic surfaces. However, there is a variety of conflicting results found in the literature concerning the antibacterial properties of su- perhydrophobic surfaces. Some groups find highly reduced bacterial adhesion,[148, 159-162] whereas other groups report enhanced adhesion.[163] Bacterial adhesion is a very complex topic and, therefore, cannot be discussed in detail herein. Many different parameters gen- erally influence the adhesion and biofilm formation, such as surface wettability, surface charge, and surface topography. Different bacteria show varied interactions with the sur- face due to differences in the composition of their cell wall and are influenced by envi- ronmental factors.[161] The differing findings for superhydrophobic surfaces may be caused by the limited stability of the Cassie state.[163] After a Cassie-to-Wenzel transition, the surface is fully wetted. Bacteria may be able to attach to the surface and form a biofilm since it has been shown that roughness can also enhance bacterial adhesion.[161, 168] Chapter 5 sheds light on the controversial question if superhydrophobic surfaces are an- tibacterial and the effect of roughness. 42 INTRODUCTION Figure 19. Super-liquid-repellent gas exchange membrane. a) Schematic illustration of gas transfer through a superamphiphobic membrane. The liquid rests on top of the superamphiphobic layer. The liq- uid-gas interface is marked in red. b) SEM image of a bare stainless steel mesh. c,d) SEM images of the ≈35 µm thick superamphiphobic coating on the steel mesh from b). The coating is based on the candle soot template. Reproduced from ref.[84] Copyright 2013, Springer Nature: Nature, licensed under CC BY- NC-SA 3.0. Few reports investigate the potential of porous super-liquid-repellent surfaces or mem- branes for gas exchange applications.[84, 169] In such a gas contactor membrane system, a liquid flows across one side of the membrane, whereas a gas streams over the other side. The membrane physically separates the liquid and gas streams.[170] The gas must be able to efficiently diffuse through the membrane. However, it is essential that the liquid does not penetrate through the membrane. The liquid is usually a solution or chemical agent that can bind one or more gases present in the gas stream. For instance, hydrogen sulfide and carbon dioxide can be bound with amine solutions[171-174], or olefins can be separated from paraffin gases using a silver nitrate solution.[175] One of the major drawbacks of such membrane systems is that the gas capture solutions or agents wet the membranes.[176] Wetting significantly decreases the performance of gas contactor membranes. It leads to a stationary liquid film inside of the membrane pores, and the gas must diffuse through this stationary film, which considerably slows down the gas exchange. Therefore, super- liquid-repellent membranes may be beneficial to prevent wetting. The capture medium stays on top of the hydrophobic protrusions leading to a high liquid-gas interface (Figure 19a). Paven et al. prepared a super-liquid-repellent membrane for oxygenation of blood.[84] A stainless steel mesh with rectangular pores of ≈30 × 30 µm2 was coated with a ≈35 µm thick super-liquid-repellent layer based on the candle soot template[41] to obtain a super-liquid-repellent membrane (Figure 19b–d). Blood on such a membrane showed ba- sically no adhesion or wetting. Even after 48 h of incubation, a blood puddle was easily rinsed off and left no traces behind. Blood showed a contact angle of 𝜃𝜃 = 162° and a roll- off angle of 𝛼𝛼 = 5° and could be oxygenated without clogging. Besides, CO2 gas exchange FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 43 using an aqueous sodium hydroxide solution was demonstrated. Despite some initial re- ports, the potential of super-liquid-repellent and superomniphobic membranes remains mostly unexplored. In Chapter 4, a superomniphobic membrane is introduced that pre- vents wetting of highly basic and concentrated amine solutions that are industrially used for CO capture[177-178]2 . This led to performance increasements of up to 40%. PARTICLES AT INTERFACES — LIQUID MARBLES There are several different scenarios for particles to adsorb at a fluid-fluid interface (Figure 20).[179] Oil-in-water emulsions are typically not stable and start demixing through the coalescence of the dispersed oil drops to reduce the overall energy of the system. Therefore, emulsions are commonly stabilized using surfactants or tensides. These am- phiphilic molecules adsorb and stabilize the liquid-liquid interface, preventing coales- cence of the oil drops. Another way to stabilize oil-in-water emulsions is through the ad- dition of hydrophilic particles. The particles adsorb at the interface of the oil drops and form a particle coating around the oil drops. These particle-stabilized emulsions are called Pickering emulsions.[180] The inverse case, a water-in-oil emulsion, can be stabilized using hydrophobic particles. In addition to liquid-liquid interfaces, also liquid-air inter- faces can be stabilized using particles resulting in air-in-water foams, dry water, and liq- uid marbles.[181-183] Figure 20. Particles at interfaces. a) Oil-in-water emulsion separated by hydrophilic particles, a so-called Pickering emulsion. b) Water-in-oil emulsion separated by hydrophobic particles, an inverse Pickering emul- sion. c) Air bubbles in water stabilized by particles at the interface, an air-in- water foam. d) Microscopic water droplets in air coated with particles, of- ten referred to as dry water. Reprinted from ref.[179] by permission from Springer Nature: Nature, Copyright 2006. The reason why particles attach and adsorb to liquid-air interfaces is to reduce the total interfacial energy (Figure 21a). The liquid-air interface forms a well-defined contact an- 44 INTRODUCTION gle 𝜃𝜃 with respect to the particle’s surface. The contact angle depends on the hydrophobi- city of the particle and determines how much the particle sticks out of the liquid-air inter- face. The distance 𝐻𝐻p the particle protrudes out of the interface is given by 𝐻𝐻p = 𝑅𝑅(1 − cos 𝜃𝜃). Here, 𝑅𝑅 is the radius of the particle, and 𝜃𝜃 is the intrinsic contact angle of the solid particle material. The adsorption energy ∆𝐺𝐺 is obtained by subtracting the solid-liq- uid interfacial energy 𝐸𝐸Interface of a spherical cap (of a particle) attached to a drop from the interfacial energy 𝐸𝐸Air, which is the interfacial energy of the same spherical cap in air and a cap-shaped portion of the liquid-air interfacial energy of the drop (Eq. 13):[184] ∆𝐺𝐺 = 𝐸𝐸Air − 𝐸𝐸Interface = 𝐴𝐴SA(𝛾𝛾SL − 𝛾𝛾SA) − 𝐴𝐴LA𝛾𝛾LA (13) Here, 𝐴𝐴SA = 𝐴𝐴SL is the solid-air and solid-liquid interfacial area of the spherical cap, 𝐴𝐴𝐿𝐿𝐿𝐿 is the liquid-air interfacial area and 𝛾𝛾SL, 𝛾𝛾SA, and 𝛾𝛾LA are the respective interfacial ener- gies. Eq. 14 follows using Young’s equation: 𝐴𝐴 ∆𝐺𝐺 = 𝐴𝐴SA𝛾𝛾LA �cos 𝜃𝜃 + LA� (14) 𝐴𝐴SA The ratio of the liquid-air interfacial area compared to the solid-air interfacial area is given by 𝐴𝐴LA⁄𝐴𝐴SA = (1 − cos𝜃𝜃)⁄2. Only for 𝜃𝜃 = 180° and 𝜃𝜃 = 0° the ratio is zero. In these cases, the particle either has no contact to the droplet (𝜃𝜃 = 180°) or enters the drop (𝜃𝜃 = 0°). For all other cases, the ratio is positive, and the attachment is energetically favored.[184] A liquid marble consists of a drop covered with a layer of particles (Figure 21b). To fabri- cate a liquid marble, a water drop is placed on a powder bed consisting of hydrophobic particles, grains, or often particle aggregates. The drop is now rolled on the powder bed. During rolling, the hydrophobic particles, grains, or aggregates attach to the drop, encap- sulating the liquid inside (Figure 21c,d). The formed liquid marble is soft, elastic, non- wetting, and can be transferred to other substrates, including hydrophilic ones. Due to the particle armor, the liquid marble does not contact and wet the underlying sub- strate.[184-188] Certain insect species, like the aphids, use these unique properties for their survival. Aphids excrete large amounts of honeydew as these insects consume nitrogen- deficient sugary liquids from the phloem of plants as a nutrient. The excreted honeydew poses a threat to their lives as entrapment results in drowning, and the dried honeydew is a perfect matrix for the growth of pathogens like fungi. Therefore, aphids living in plant galls produce wax tufts to coat the inside of their gall. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 45 Figure 21. Liquid Marbles. a) Particles adsorb at the air-water interface to minimize interfacial energy. b) Droplet coated with particles on a solid substrate, a liquid marble. The particle layer separates the wa- ter drop from the underlying substrate. c) Formation of a liquid marble by rolling a water drop over a hy- drophobic powder consisting of hydrophobic grains. During rolling, the hydrophobic grains attach to the drop. d) Photograph of a liquid marble. Reproduced from ref.[184] with permission from The Royal Society of Chemistry. e) Liquid marbles in nature. The armored droplets consist of honeydew encapsulated by wax powder, which was produced by aphids. Scale bar: 1 mm. Reproduced with permission from ref.[189] Copyright 2002, The Royal Society. The wax consists of long-chained alcohols, aldehydes, esters, and fatty acids and is highly hydrophobic. The excreted honeydew is now encapsulated by the wax powders leading to the formation of liquid marbles (Figure 21e). These honeydew liquid marbles are non- wetting and can be moved by the aphids. Soldier aphids do not only defend the gall but are also responsible for cleaning, i.e., rolling the honeydew liquid marbles out of the gall through a small opening, ensuring the survival of the aphids.[189] Liquid marbles have been prepared using various kinds of stabilizers, including organic,[190-192] inorganic,[193-195] and composite[196] particles and powders. They can encapsulate different liquids, including functional liquids like blood.[197-198] Xue et al. prepared F-POSS powders and F-POSS/Fe3O4 nanoparticle composite powders to encapsulate various kinds of polar and non-polar liq- uids, including water, dimethyl sulfoxide, toluene, ethanol, and octane (Figure 22a). 46 INTRODUCTION Figure 22. Liquid marbles prepared with F-POSS powder. a) Various liquids (3 µL), including water, di- methyl sulfoxide (DMSO), toluene, ethanol, and octane, can be encapsulated using the hydrophobic fluori- nated powder. Dyes were added to the liquids for better visibility. b) Liquid marbles prepared with F-POSS powder and various liquids (3 µL) floating on water and hexadecane. Dyes were added for visibility. Re- produced with permission from ref.[199] Copyright 2010, John Wiley and Sons. These liquids have largely different surface tensions, even down to 𝛾𝛾LG = 21.6 mN m-1 for octane. The liquid marbles could be transferred and float on a water surface and even on a hexadecane surface (Figure 22b). Liquid marbles prepared with the F-POSS/Fe3O4 nano- particle composite powder could be opened and moved using magnetic fields and be uti- lized as a reactor. Besides that, liquid marbles were used in diverse applications, includ- ing transport of liquids,[200] miniature reactors,[194, 199] micro bioreactors, and drug screen- ing,[198, 201-202] microcapsules,[192] sensors,[190] accelerometers,[203] and pressure-sensitive ad- hesives.[193] For instance, Fujii et al. embedded pressure-sensitive adhesive (PSA) polymer solutions in liquid marbles (Figure 23). As a liquid marble stabilizer, CaCO3 nanoparticles were hydrophobized using octadecanoic acid. Poly(n-butyl acrylate) (PBA) latex particles were synthesized as PSA base polymer. An aqueous dispersion (180 µL) of the PBA latex particles (≈50 wt% solid content) was deposited on the hydrophobized CaCO3 powder, which immediately encapsulated the PBA particle dispersion. Afterward, the formed liq- uid marbles were dried for several hours to evaporate the water content. The dried PSA marbles remained non-sticky and were flowing freely like a powder. After the application of shear stress, the pressure-sensitive adhesive could be released unfolding its adhesive behavior. These PSA marbles may be useful to transport glues into confined, hard-to- reach spaces.[193] FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 47 Figure 23. Pressure-sensitive adhesive (PSA) Marbles. a) A PSA is armored with hydrophobic CaCO3 nanoparticle powder. After encapsulation, the PSA marbles are non-sticky and act as a free-flowing pow- der. b) Only upon applying pressure, the PSA marbles break and release the sticky PSA polymer. Reproduced from ref.[193] with permission from The Royal Society of Chemistry. It should be noted that there is a distinct similarity between liquid marbles and superhy- drophobic or superamphiphobic surfaces (Figure 24). A water droplet on a superhydro- phobic surface takes an almost spherical shape due to the rough and hydrophobic layer on the substrate. On the other hand, a liquid marble is a droplet coated with a rough, hy- drophobic layer. Thus, a liquid marble can be regarded as a droplet covered by a super- hydrophobic layer that was previously detached from the substrate.[187] Figure 24. Similarities between superhy- drophobic surfaces and liquid Marbles. a) Superhydrophobic layer attached to a substrate. b) Superhydrophobic layer on a liquid marble. Reproduced from ref.[187] with permission from The Royal Society of Chemistry. The in-situ formation of a liquid marble is observed during the self-cleaning of a superhy- drophobic surface in Chapter 3, and these principles are used in Chapter 6 to prepare unique liquid marbles. 48 HOW DROPS START SLIDING OVER SOLID SURFACES 2 HOW DROPS START SLIDING OVER SOLID SURFACES Nan Gao,1,2 Florian Geyer,1 Dominik W. Pilat,1 Sanghyuk Wooh,1 Doris Vollmer,1 Hans- Jürgen Butt,1 and Rüdiger Berger1 1Max Planck Institute for Polymer Research, Ackermannweg 10, 55128 Mainz, Germany 2Fudan University, 220 Handan Road, Shanghai 200433, People’s Republic of China Abstract: It has been known for more than 200 years that the maximum static friction force between two solid surfaces is usually greater than the kinetic friction force — the force that is required to maintain the relative motion of the surfaces once the static force has been overcome. But the forces that impede the lateral motion of a drop of liquid on a solid surface are not as well characterized, and there is a lack of understanding about liquid-solid friction in general. Here, we report that the lateral adhesion force between a liquid drop and a solid can also be divided into a static and a kinetic regime. This striking analogy with solid-solid friction is a generic phenomenon that holds for liquids of differ- ent polarities and surface tensions on smooth, rough, and structured surfaces. Keywords: Applied physics, fluids This chapter is reprinted with permission from N. Gao, F. Geyer et al., Nat. Phys. 2018, 14, 191. Copyright 2018, Springer Nature: Nature The Supporting Information is available from nature.com: https://www.nature.com/articles/nphys4305#supplementary-information FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 49 INTRODUCTION When two solid objects are brought into contact, a threshold force FTHRD must be over- come in order for one of the objects to slide.[204-206] This phenomenon can be visualized in a typical classroom experiment where a solid block attached to a spring is pulled over a solid surface (Figure 25a). The static force 𝐹𝐹S is applied to the stationary block and then increased until it exceeds 𝐹𝐹THRD, upon which the block begins to slide. After that, a lower kinetic force 𝐹𝐹KIN is required to maintain the block’s motion.[206] However, it is not clear whether these forces develop in a comparable manner when a drop of liquid resting on a solid surface starts to slide. This gap in our understanding is astonishing, given the fact that liquid drops are omnipresent in our lives, and their motion is relevant for numerous applications, including microfluidics,[11] printing,[207] condensation,[14-15] and water collec- tion.[12-13] Hence insight on the behavior of drops that start sliding over solid surfaces is needed. A sessile drop of liquid is usually in molecular contact with the supporting solid surface. In contrast, two solid bodies are in direct contact only at asperities owing to surface rough- ness.[208-209] Thus, the real contact area of a solid-solid contact is much smaller than the apparent contact area. Consequently, the sliding of drops might be fundamentally differ- ent. However, by simply observing a drop of water on a pivot window pane, we know that also sessile drops start sliding when a critical tilt angle is reached, i.e., when the gravita- tional force acting on the drop overcomes the lateral adhesion force.[210] The question may, therefore, be raised whether a static and a kinetic regime are also present for sessile drops. The general question is: How do drops start sliding over solid surfaces, and how do the forces develop while the drops slide? Owing to higher gravitational forces, larger drops start sliding at lower tilt angles. Sliding is opposed by capillary forces. They are associated with a contact angle difference be- tween the rear and the front of the drop. Indeed, the interactions between solid surfaces and liquids are described by the liquid-air surface tension 𝛾𝛾 and the apparent rear and front contact angles of the drop, 𝜃𝜃Rear and 𝜃𝜃Front, respectively. Thus, the surface tension, the contact angles and the drop contact width L determine the lateral adhesion force 𝐹𝐹LA by[211-215] ∆𝐹𝐹LA = 𝑘𝑘𝑘𝑘𝛾𝛾(cos𝜃𝜃Rear − cos 𝜃𝜃Front). (15) 50 HOW DROPS START SLIDING OVER SOLID SURFACES The dimensionless factor k accounts for the precise shape of the solid-liquid-air three- phase contact line of the drop. Values for k were calculated to be between 1/2 and π/2.[211, 216-218] Despite the omnipresence of drops, the onset of motion has never been correlated with the development of lateral adhesion forces. The lateral adhesion force has been related to external forces that cause a drop to slide, such as gravitational,[219-220] centrifugal,[9] mag- netic,[221] or capillary forces.[6-8] The contact angles have also been experimentally and nu- merically investigated for the pinned state, i.e., just before and during steady mo- tions.[222-223] However, once a drop has started to slide, the lateral adhesion force cannot be tracked using simple techniques. Astonishingly, it is unclear how the force develops and how it depends on sliding velocity. We will demonstrate that, for the lateral liquid- solid adhesion, we can distinguish a static and a kinetic regime, analogous to solid-solid friction. RESULTS AND DISCUSSION To measure the lateral adhesion force between a drop of liquid and a solid substrate, a capillary is positioned in the center of the drop. The substrate with the drop is moved sideways against the capillary at a constant velocity. When the capillary reaches the edge of the drop, it sticks to the drop. Consequently, the motion of the substrate is accompa- nied by a deformation of the drop as well as a deflection of the capillary. Initially, the drop remains pinned to the substrate (Figure 25b). Figure 25. Schematics of friction force measurements. a) Textbook configuration for demonstrating solid-solid friction. b) Homemade setup for measuring liquid-solid friction. A drop of liquid is placed on a solid substrate mounted on a linear stage driven by a step motor. A laser beam incident on the capillary is reflected to a position-sensitive detector (PSD). The contact width between the drop of liquid and the solid surface (orthogonal to the direction of motion) and contact length (parallel to the direction of mo- tion) are simultaneously monitored by cameras (not shown). FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 51 Once the capillary exerts a certain critical force, the drop overcomes the lateral adhesion and is set into translational motion relative to the substrate, i.e., the front and rear side of the drop start moving. The deflection 𝐷𝐷 of the capillary is measured by recording the po- sition of a reflected laser beam with a position-sensitive detector.[7] Then, the lateral ad- hesion force acting on the drop can be calculated by 𝐹𝐹LA = 𝜅𝜅𝐷𝐷, where 𝜅𝜅 is the spring con- stant of the capillary. Simultaneously, the drop’s shape is monitored by two cameras, which are synchronized with the force measurement. In this way, the lateral adhesion force can be correlated with the contact angles. Velocities ranged from 1 µm s-1 to 50 mm s-1 (see Experimental Section 2.3). As a representative example, we start with a drop of an ionic liquid placed on a fluorinated Si wafer. It forms a contact angle of ≈70°. Moving the wafer laterally increased the force (blue circles in Figure 26a) until a maximum force of 50 µN was reached (after 11 s). This maximum force corresponded to the threshold force for the drop, upon which it started to slide. Then the force subsequently decreased to a constant level of ≈33 µN, which was only 66% of the maximal force (after 17 s). Accordingly, we distinguish a static, a transi- tion, and a kinetic regime shown in different shades of green in Figure 26a. The presence of a higher threshold force compared to the force required for continuous motion ex- plains the manner a drop slides downhill in a tilted-plate experiment. It is analogous to solid-solid friction, where a solid object placed on a tilted plane keeps sliding once the static friction threshold force has been overcome.[224] Initially, the contact line of the drop was pinned, and the contact width and length remained constant at about 2.3 mm (Figure 26b, first 5 s). Meanwhile, the front and rear contact angles increased and de- creased, respectively, as the drop deformed (Figure 26c). After 5–8 s, the contact line started reforming, resulting in a longer and narrower drop shape. After about 10 s, the contact angles reached a maximum value of ≈86° at the front and a minimum of ≈43° at the rear. At this point of maximum contact angle difference of ≈40°, the static adhesion force reached its maximum, which constituted the threshold force of the translational motion. In the transition regime, the contact length decreased continuously. The rear contact angle increased by more than 10°, whereas the front contact angle decreased by ≈5° over time. Finally, a kinetic regime was reached (after ≈18 s), which was characterized by constant lateral adhesion forces, constant contact angles, constant contact widths, and contact lengths. 52 HOW DROPS START SLIDING OVER SOLID SURFACES Figure 26. Lateral adhesion force experiment of a drop of ionic liquid (volume ≈1.5 µL) on a fluorinated silicon wafer. The measurement was performed at a constant linear stage velocity of 200 μm s-1. The surface tension of the ionic liquid 1-butyl- 2,3-dimethylimidazolium bis(trifluorometh- anesulfonyl)imide is 34.6 mN m-1. a) The measured lateral adhesion force is plotted with blue circles. The error in force measurement is ±1 µN and smaller than the symbol size. The calculated force is plotted with red Squares. b) Simultaneously deter- mined contact width and length of the drop during the force measurement shown in a). c) Front and rear contact angles during the force measurement. Movie S1 shows the motion of the ionic liquid drop on the fluorinated Si wafer during the lateral adhesion force measurement. We then took the measured contact widths (Figure 26b) and contact angles (Figure 26c) to calculate the lateral adhesion force using Eq. 1. Thereby we simplified Eq. 1 by assum- ing a constant value of 𝑘𝑘 = 1 thus disregarding shape changes of the drop contour and variations of the contact angle along the contour[216] (see Supplementary Information for a discussion of the 𝑘𝑘 factor and its temporal development). This calculation almost quan- titatively agreed with the measured lateral adhesion force (red squares in Figure 26a). In particular, the calculated lateral adhesion force reproduced the transition between the static and kinetic regimes. Once the threshold adhesion force 𝐹𝐹THRD has been overcome, the drop transitions from a static regime to a steady kinetic regime. Consequently, the liquid-solid adhesion forces in the static and kinetic regimes are substantially different, owing to different contact angles and contact widths in these two regimes: FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 53 𝐹𝐹 RearS = 𝛾𝛾𝑘𝑘S�cos𝜃𝜃S − cos 𝜃𝜃FrontS � (16) 𝐹𝐹 Rear FrontKIN = 𝛾𝛾𝑘𝑘KIN�cos𝜃𝜃KIN − cos 𝜃𝜃KIN � (17) Here, 𝐹𝐹 , 𝑘𝑘 , 𝜃𝜃Front and 𝜃𝜃RearS S S S are the lateral adhesion force, contact width, and front and rear contact angles in the static (S) regime. 𝐹𝐹 , 𝑘𝑘 Front RearKIN KIN, 𝜃𝜃KIN and 𝜃𝜃KIN are the lateral ad- hesion force, contact width, and front and rear contact angles in the kinetic (KIN) regime. To verify whether the discrimination between static and kinetic lateral adhesion forces is a generic feature, we analyzed a number of different liquid-solid combinations (Figure 27). We used water and hexadecane in addition to a non-volatile ionic liquid (1-bu- tyl-2,3-dimethylimidazolium bis(trifluoromethanesulfonyl)imide). The solid surfaces in- clude superhydrophobic fluorinated silicone nanofilaments (SNFs), fluorinated silicon wafers (Si), fluorinated SU-8 micropillar arrays, fluorinated multilayers of titanium diox- ide nanoparticles (TiO2) and cross-linked polydimethylsiloxane (PDMS) substrates (Figure S1). All samples were homogeneous at the scale of the drop radius and did not have distinct pinning centers. For all liquid-solid combinations, we observed a static, a transition, and a kinetic regime (Figure 27a, Figure S5–S10). It should be noted that the lateral adhesion force in the kinetic regime never exceeded the threshold force (Figure 27a): 𝐹𝐹THRD ≥ 𝐹𝐹KIN (18) Notably, 𝐹𝐹THRD = 𝐹𝐹KIN is possible, too.[7] Surfaces coated with brushes of liquid-like PDMS showed no difference between 𝐹𝐹THRD and 𝐹𝐹KIN. The reason is that these surfaces show al- most identical contact angle hysteresis for water at the threshold and in the kinetic re- gimes (Figure S11). 54 HOW DROPS START SLIDING OVER SOLID SURFACES Figure 27. Lateral adhesion forces for drops of different liquids on solid surfaces. a) Measured (blue circles) and calculated lateral adhesion forces using k = 1 (red squares). The surfaces include silicone nanofilaments (SNFs), silicon wafers (Si), SU-8 square pillar arrays (height: 25 µm, width: 50 µm, cen- ter-center distance: 100 µm), multilayers of 20-nm titanium dioxide nanoparticles (TiO2 NPs), and cross- linked polydimethylsiloxane (PDMS). All surfaces except for those of PDMS were fluorinated before use. All fabrication details are provided in the Experimental Section 2.3. The adopted surface tensions of water and hexadecane are 72.5 mN m-1 and 27.5 mN m-1, respectively. The front and rear contact angles in the kinetic regimes were 171° ± 1° and 164 ± 1° for water on fluorinated silicone nanofilaments, 83° ± 1° and 59° ± 1° for hexadecane on fluorinated silicon wafers, 169° ± 2° and 122° ± 2° for water on fluorinated SU-8 pillars, 128° ± 1° and 95° ± 1° for water on fluorinated silicon wafers, 164° ± 1° and 128° ± 2° for water on titanium dioxide nanoparticles, and 121° ± 1° and 81° ± 1° for water on crosslinked PDMS, respectively (Figure S5–S10). Movies S2–S7 show the motions of drops on the different surfaces during the lateral ad- hesion force measurements. b) Lateral adhesion force per contact width. Drop volumes between 1.5 and 8.0 µL were chosen to avoid the rupturing of drops during motion. c) Ratios of the kinetic friction force FKIN and its threshold force FTHRD of all liquid-solid combinations that are studied. Here we used FKIN and FTHRD prior normalization with the respective contact widths in order to point out FKIN/FTHRD ≤ 1. Error bars in b and c indicate variability between different experiments. Furthermore, good agreement between calculated and measured lateral adhesion forces was observed for various wetting scenarios, including hexadecane and ionic liquid on FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 55 fluorinated silicon wafers, and water on superhydrophobic silicone nanofilaments and fluorinated SU-8 pillars, where drops partially rested on air cushions. The differences in adhesion between the drops and the substrates resulted in different apparent contact an- gles and contact widths among the liquid-solid combinations (Figure S5–S10). We attrib- ute the small discrepancies between calculated and measured lateral adhesion forces to microscopic distortions of the three-phase contact lines.[210] The latter can be accounted for by the 𝑘𝑘 factor (Figure S5–S10 and Table S1). Indeed, the 𝑘𝑘 factor of our measurements lies between 0.25 and 4. It is worth emphasizing that the 𝑘𝑘 factor changed continuously while the drop passed from the static to the kinetic regime. These observations indicate that a constant 𝑘𝑘 factor is not sufficient to describe how drops slide over a solid surface. For large scale heterogeneous surfaces that exhibit distinct pinning centers, 𝐹𝐹KIN occa- sionally exceeded 𝐹𝐹THRD (Figure S12 and Griffiths and colleagues[225]). Thus, the lateral ad- hesion force provides quantitative information on the homogeneity or cleanliness of a surface. To compare the different liquid-solid combinations independently from drop size, we cal- culated the static and lateral adhesion force per contact width (Figure 27b). This analysis yielded values ranging from 2 µN mm-1 to 110 µN mm-1. The increasing lateral adhesion force per contact width depends on the interplay between surface tension and contact angle hysteresis. In addition, we calculated the ratios of the kinetic force divided by the static threshold force. For all different liquid-solid combinations, we obtained a ratio of 𝐹𝐹KIN/𝐹𝐹THRD ≤ 1 (Figure 27c). In general, contact angles depend on the sliding velocity.[226] Therefore we measured 𝐹𝐹KIN and 𝐹𝐹THRD in a velocity range from 10-3 to 34 mm s-1 for hexadecane on a fluorinated silicon wafer surface (filled symbols in Figure 28a,b). We found that 𝐹𝐹KIN and 𝐹𝐹THRD are relatively constant up to a capillary number Ca of 10-5. Then they increase with velocity. Further- more, the dependence and magnitude of the measured forces are in agreement with the forces calculated by the front and rear contact angles (inset of Figure 28b) using Eq. 15 (open symbols in Figure 28a,b). Similarly, an increase in lateral adhesion forces was ob- served for ionic liquid and water on fluorinated silicon wafers, respectively (Figure 28c,d). Perrin et al. measured the dynamic macroscopic contact angle of a silicon wafer with- drawn from a silicone oil bath.[227] These experiments revealed a strong change in the co- sines of the contact angles between 10-5 < Ca < 10-3. This region marked the transition from energy dissipation dominated by defects to viscous dissipation. 56 HOW DROPS START SLIDING OVER SOLID SURFACES Figure 28. a,b) Velocity dependence of lateral adhesion forces. Development of the lateral adhesion force of hexadecane drops (≈3 μL) on a fluorinated Si wafer surface (a) at the threshold and (b) in the ki- netic regime. The inset shows the front and rear contact angles measured in the kinetic regime. c,d) Velocity dependence of lateral adhesion forces of an ionic liquid (1-butyl-2,3-dimethylimidazolium bis(trifluoromethanesulfonyl)imide) drop (≈2 μL) on a fluorinated Si wafer substrate (c) and water drops (≈7.5 μL) on a fluorinated Si wafer substrate (d). Here velocities below 0.08 mm s-1 were not studied owing to evaporation of water. The error bars correspond to the standard deviation calculated from 5–7 inde- pendent measurements. The highest velocity that can be probed is given by the threshold force where the liquid drop detaches from the glass capillary (≈81 µN for water, ≈36 µN for ionic liquid, and ≈31 µN for hexadecane). The increase in 𝐹𝐹KIN falls in this regime, and therefore we attribute the increase of 𝐹𝐹KIN at Ca > 10-5 to viscous dissipation in addition to thermally activated processes at the contact line. The latter was discussed as a result of a rugged energy landscape of the sample sur- face, leading to a thermally activated process of the contact line dynamics.[227] In our meas- urements, we could not resolve a logarithmic dependence of 𝐹𝐹 -5KIN and 𝐹𝐹THRD at Ca < 10 , which would have given insight into such a thermally activated process of the moving liquid. In particular, we found that the ratio of 𝐹𝐹KIN/𝐹𝐹THRD ≤ 1 holds within the probed velocity range (Figure S13 and S14). FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 57 Figure 29. Lateral adhesion force measure- ment of a water drop on a goose feather. a) Three different lateral adhesion force measurements of a water drop on a goose feather at different areas at a velocity of 6.8 mm s-1 (drop volume of 8.0 µL). The threshold lateral adhesion force varied be- tween 35 ± 1 µN and 67 ± 2 µN. Continuing the kinetic motion of the water drop re- quired a force up to 35 µN. b) An optical im- age of the feather with resting water drops. c) Scanning electron microscopy images re- veal the detailed structure of the feather. The inset is an enlarged area showing the microscopic structure. The scale bar in the inset is 10 µm. d) No velocity dependence of the lateral adhesion force at threshold (black squares) and in the kinetic regime (red circles) could be measured. The error bars correspond to the standard deviations calculated from 5–7 measurements each. In order to study whether drop-solid friction scales with the contact line or with the con- tact area, we varied the drop volume resulting in a variation of the contact area by a factor of >4 (Figure S13). The linear extrapolation of the kinetic force suggests a linear depend- ence with an intercept slightly below zero force. Therefore we conclude that the liquid drop friction is dominated by contact line friction, and interfacial friction only plays a minor role. This conclusion is supported by an estimation of the friction force contribu- tion of a solid-liquid interface based on a Green-Kubo relation of Huang and Szlufarska[228] 58 HOW DROPS START SLIDING OVER SOLID SURFACES and by monitoring the motion of drops over the surfaces, showing that the motion is dom- inated by rolling (Figure S15, Movies S8–S11). In contrast, contact line friction does not play a role for liquid-liquid interfaces, such as drops on lubricated surfaces.[229] As an example of a natural, structured surface, we measured the lateral adhesion of a water drop on a goose feather (Figure 29a). Again, we observed a static regime and a ki- netic regime. The static threshold force was up to 3 times larger than the average kinetic force required for the continuous motion of the water drop. In addition, the lateral adhe- sion force varied in the kinetic regime. Likely, these variations are related to the flexibility and the microscopic structures of the feather (Figure 29b,c), which lead to a stick-slip motion of the drop. Energy is dissipated mainly by depinning from a few surface protru- sions, which shows no velocity dependence of the lateral adhesion force at the threshold and in the kinetic regime (Figure 29d and Figure S16).[230] The latter is in contrast to the stick-slip motion of two solids.[231-232] Consequently, for such rough and elastic surfaces, the contribution of viscous dissipation to the overall energy dissipation of the sliding drop is low. It is not clear what role such a stick-slip motion and the associated velocity inde- pendence play for the life of a goose in water. The constant and low kinetic force could potentially reduce energy losses at the three-phase contact line of the feather with water and air. EXPERIMENTAL SECTION MATERIALS The following chemicals were used to fabricate the solid substrates: trichloro- methylsilane (TCMS, 99%, Sigma-Aldrich), 1H,1H,2H,2H-perfluorodecyltrichlorosilane (PFDTS, 96%, Alfa Aesar), 1H,1H,2H,2H-perfluorooctyltrichlorosilane (PFOTS, 97%, Sigma-Aldrich), trimethylsiloxy-terminated polydimethylsiloxane (PDMS, 200 cSt, ABCR), sulfuric acid (95%, VWR), Sylgard 184 (PDMS precursor, Dow Corning), Sylgard 184 curing agent (Dow Corning), ammonia solution (28%, VWR), hydrogen peroxide solution (Sigma- Aldrich, 34.5−36.5%), n-hexane (99.99%, Fisher Chemical), toluene (99.99%, Fisher Chemical), isopropanol (99.5%, Fisher Chemical), acetone (99.98%, Fisher Chemical), ethanol (absolute, 99.96%, VWR Chemicals), tetrahydrofuran (99.9%, Sigma-Aldrich), SU-8 3025 photoresist (MicroChem), mr-Dev 600 developer (micro resist technology), 18NR-T (Dyesol) and sodium hydroxide (98%, Sigma-Aldrich). Reagents were used as received. Polished silicon wafers were obtained from Si-Mat. Thin glass slides of FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 59 24 × 60 mm2 and 170 ± 5 µm thickness were obtained from Carl-Roth. All syntheses were carried out at room temperature. The liquids drops used for force measurements included those of hexadecane, 1-butyl-2,3-dimethylimidazolium bis(trifluoromethanesulfonyl)- imide and water. Hexadecane (99%, Sigma-Aldrich) and 1-butyl-2,3-dimethylimidazolium bis(trifluoromethanesulfonyl)imide (99%, IoLiTec) were used as received. Water with a typical resistivity of 18.2 MΩ·cm was obtained from a Sartorius Arium 661 VF Water Purification System. SILICONE NANOFILAMENTS Silicone nanofilaments on glass slides were synthesized by modifying previously reported methods.[112, 115] The fabrication contained the following steps:[116, 118, 233] First, 120 µL of TCMS was added into a reaction chamber containing 50 mL of water-saturated n-hexane. The resulting solution was stirred for 60 s. Then, cover glass slides were immersed in the solution before sealing the reaction chamber. After 3 days, the TCMS-coated glass slides were removed from the reaction chamber, rinsed with n-hexane, and dried under a nitro- gen stream. The TCMS-coated glass slides were activated in an oxygen plasma chamber (75 W, 2 min, 7 cm3 min-1 oxygen flow rate, Diener Electronic Femto) before modification with the fluorinated silane PFDTS. Therefore, the activated substrates were immersed in a solution containing 25 µL of PFDTS dispersed in 50 mL of n-hexane for 20 min. Subse- quently, the substrates were rinsed with n-hexane and dried under a nitrogen stream. Representative scanning electron microscopic images of the resulting silicone nanofila- ments are shown in Figure S1a,b. SILICON WAFERS Pristine silicon wafers were cleaned by ultra-sonication in toluene, acetone, and ethanol for 15 min, respectively. Subsequently, the pre-cleaned wafers were plasma-cleaned and activated in an oxygen plasma chamber (300 W, 10 min, 6 cm3 min-1 oxygen flow rate). To fluorinate the surfaces, 25 µL of PFDTS was mixed with 50 mL of n-hexane. The activated silicon wafers were immersed in the solution for 5 to 15 min. Afterward, the silicon wafers were rinsed with n-hexane, followed by ultrasonication in isopropanol for 5 min. The cleaned silicon wafers were dried under a nitrogen stream. It should be noted that the immersion time influences the lateral adhesion force of the fluorinated silicon wafers. To access higher sliding velocities, we improved the preparation of the Si wafers. We found that ultrasonication in isopropanol helps to remove residues from the fluorination. Gen- erally, shorter fluorination times (5 min) yielded a lower adhesion. Therefore the lateral 60 HOW DROPS START SLIDING OVER SOLID SURFACES adhesion forces measured for fluorinated silicon wafers in Figure 28 are lower compared to Figure 26. Lower lateral adhesion forces allowed us to investigate higher velocities. The scanning electron microscopy analysis revealed featureless, smooth surfaces, as shown in Figure S1c,d. SU-8 PILLARS SU-8 micropillar arrays were prepared on thin glass slides by photolithography.[62] The square pillars were designed 25 µm high, with 50 × 50 µm2 top areas. The pillar–pillar dis- tance between centers of two adjacent pillars in a row was 100 µm. The fabrication process consisted of the following steps. First, glass slides (Carl-Roth) were cleaned by ultra-son- ication in tetrahydrofuran, acetone, and ethanol for 15 min, respectively. Then, SU-8 pho- toresist was spin-coated (500 rpm for 5 s followed by 3000 rpm for 30 s, SÜSS MicroTec) on the glass slides. The coated glass slides were heated at 65 °C for 3 min, 95 °C for 10 min, and then 65 °C for 30 min, respectively. Subsequently, the samples were slowly cooled down within 2 h and exposed to UV light (mercury lamp, 350 W) under a photolithography mask for 15 s (masker aligner SÜSS MicroTec MJB3 UV400). To crosslink the photoresist, the samples were heated up again at 65 °C for 3 min, 95° C for 10 min, and 65 °C for 30 min, and then cooled down slowly. Next, the samples were immersed in the SU-8 developer solution for 5 min, washed with isopropanol, and then dried in air, resulting in surfaces coated with SU-8 pillars. The dried samples were immersed in a 0.1 M NaOH solution overnight to hydrolyze the surfaces of the SU-8 pillars. After rinsing the hydrolyzed sur- faces with water and ethanol, the samples were immersed in a solution containing 25 µL of PFDTS dispersed in 50 mL of n-hexane for 20 min. Finally, the fluorinated SU-8 mi- cropillar surfaces were rinsed with n-hexane and dried under a nitrogen stream. Repre- sentative scanning electron microscopic images of an obtained SU-8 micropillar array are shown in Figure S1e,f. PDMS Crosslinked PDMS surfaces: Sylgard 184 PDMS precursor was mixed with Sylgard 184 cur- ing agent in a ratio of 10:1 by weight and then degassed three times in a vacuum chamber for 5 min, respectively. Afterward, the viscous mixture was deposited on glass slides and cured at 60 °C overnight. The scanning electron microscopy analysis revealed featureless, smooth surfaces, as shown in Figure S1g,h. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 61 Liquid-like PDMS surfaces: So termed “liquid-like” PDMS samples were prepared follow- ing the method reported by Krumpfer and McCarthy.[234] Pre-cleaned cover slides were placed in a 2:1 mixture of sulfuric acid and hydrogen peroxide solution, placed in a water bath and held at 70 °C for 2 h. Then the samples were rinsed with pure water and ethanol. After being dried, the cover slides were put into 500 mL ultraclean borosilicate glass vials (Duran Pure vials GL45 with a dust cover, Duran Group, SCHOTT) and sufficiently wetted with PDMS. The vials were resealed and placed in an oven at 100 °C for 24 h. Then the samples were properly rinsed with n-hexane, acetone, and water and dried. The flexibility of the O-Si-O bonds confers high mobility to PDMS chains. As only one end of the PDMS is covalently grafted on the substrate, the remaining part of the PDMS keeps its high mobility with rotational and/or bending motion. Thus, these grafted layers are rotationally dynamic and behave “liquid-like.” Water drops slide off these “liquid-like” surfaces when tilting the surface by 4° (drop volume: 20 mL).[234] TITANIUM DIOXIDE NANOPARTICLES TiO2 surfaces consisting of ≈20 nm TiO2 nanoparticles (NPs) were fabricated from TiO2 nanoparticle paste (18NR-T, Dyesol).[235] First, TiO2 nanoparticle paste was coated on a sil- icon wafer by doctor blade coating with 50 µm spacers. Then, the coated wafer was sin- tered at 500 °C for 30 min, resulting in nano-roughened TiO2 structures. After cooling down, the TiO2 structures were modified with PFOTS by chemical vapor deposition (CVD) in a desiccator for 2 h in vacuo. Scanning electron microscopic images of a resulting TiO2 surface are shown in Figure S1i,j. FORCE MEASUREMENTS: SENSITIVITY OF THE LASER DEFLECTION SYSTEM For the measurement of the lateral adhesion force by means of our laser deflection sys- tem illustrated in Figure 25b, a hollow rectangle capillary was used to move a liquid drop against a substrate underneath.[7, 236] A laser beam was incident on the capillary, which reflected the beam to the center of a position-sensitive detector (PSD). The capillary was manufactured in borosilicate glass (VitroCom's Vitrotubes, CM Scientific). It was 35 mm long. The dimensions of the capillary were 0.04 × 0.40 mm2, with a wall thickness of 0.028 mm. In order to increase the reflectivity of the capillary, its wide side was sputter- coated with a 5-nm-thick layer of chromium and a 50-nm-thick layer of gold. The chro- mium was used as an adhesion layer for the gold. When the laser hit the PSD, four photo- currents were generated by the PSD at two opposite sides along the horizontal axis and two opposite sides along the vertical axis. These four photocurrents were collected and 62 HOW DROPS START SLIDING OVER SOLID SURFACES converted into four voltages (𝑈𝑈𝑥𝑥1, 𝑈𝑈𝑥𝑥2, 𝑈𝑈𝑦𝑦1, and 𝑈𝑈𝑦𝑦2) using an amplifier with a current-to- voltage converter. These four voltages were read out in two values by the amplifier corre- sponding to the horizontal (𝑥𝑥) and vertical (𝑦𝑦) coordinates of the laser spot on the PSD: 𝑈𝑈 − 𝑈𝑈 𝑈𝑈𝑥𝑥 = 5V ∙ 𝑥𝑥1 𝑥𝑥2 (19) 𝑈𝑈𝑥𝑥1 + 𝑈𝑈𝑥𝑥2 𝑈𝑈𝑦𝑦1 − 𝑈𝑈𝑈𝑈 𝑦𝑦2𝑦𝑦 = 5V ∙ 𝑈𝑈 + 𝑈𝑈 (20) 𝑦𝑦1 𝑦𝑦2 The divisions were performed by an analog divider (Burr-Brown DIV100). 𝑈𝑈𝑦𝑦 = 𝑈𝑈PSD and corresponded to the deflection of the capillary in the measurement of the lateral adhesion force. These voltage values were digitalized in an AD converter (National Instruments (NI)). Furthermore, the data acquisition software-generated triggers for the cameras to ensure synchronization between the cameras and the PSD signals. The relationship between the deflection of the capillary, 𝐷𝐷, and the generated 𝑈𝑈PSD rep- resents the sensitivity 𝑆𝑆 of the laser deflection system: 𝑈𝑈 𝑆𝑆 = PSD (21) 𝐷𝐷 For calibration, a defined deflection of the capillary was applied close to its free end by a micromanipulator. A linear correlation was found between 𝑈𝑈PSD and 𝐷𝐷 for deflections up to 800 µm. The corresponding sensitivities were in the range of -4 to -5 × 10-3 V µm-1. We performed a calibration for each new capillary and after each readjustment of the setup. Figure S2 shows a representative plot, with 𝑆𝑆 = -4.16 × 10-3 V µm-1 and Pearson 𝑟𝑟 = 0.9999. The sign of the voltage corresponds to the direction of the laser displacement on the PSD along the y-axis. SPRING CONSTANT OF THE CAPILLARY The spring constant κ of the glass capillary was calibrated by measuring the force acting on the capillary with respect to the lateral deflection. First, the capillary was fixed at one end into a copper holder as a fixture (Figure S3). Then the entire fixture was mounted to a micromanipulator (MMO-203, Narishige), which allowed us to move the fixture pre- cisely (with 1 µm precision). A pin with supporting pedestals at the bottom and a pointy tip facing up were placed on a microgram balance (Figure S3). Using the micromanipula- tor, the free end of the capillary was located right above the tip of the pin. Afterward, the capillary was lowered down slowly by means of the micromanipulator to engage with the FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 63 tip until a slight deflection of the capillary was generated. The place where the free end of the capillary was contacted by the pin was within 1 ± 0.2 mm from the edge of the free end. The microgram balance read out the weight that reflected the force, F, between the pin and the deflected capillary, as 𝐹𝐹 = 𝑚𝑚𝑚𝑚, where 𝑚𝑚 = 9.81 m s-2. Figure S4 is a representative plot of the calibration of the spring constant of the glass capillary. The relation between the force, 𝐹𝐹, and the deflection, 𝐷𝐷, followed Hooke’s law[237] 𝐹𝐹 = 𝜅𝜅 ∙ 𝐷𝐷 (22) with 𝜅𝜅 = 0.202 ± 0.002 N m-1. FORCE CONVERSION In order to obtain the lateral adhesion force, the voltage values directly collected from the PSD need to be converted into forces. Combining Eq. 21 and Eq. 22 yielded the relation- ship between the lateral adhesion forces, 𝐹𝐹LA, and the generated voltage: 𝑈𝑈 𝑈𝑈 𝐹𝐹LA ≡ 𝐹𝐹 = 𝜅𝜅 ∙ 𝐷𝐷 = 𝜅𝜅 ∙ PSD 𝐹𝐹LA ≡ 𝐹𝐹 = 𝜅𝜅 ∙ 𝐷𝐷 = 𝜅𝜅 ∙ PSD (23) 𝑆𝑆 𝑆𝑆 SUBSTRATE MOVEMENT SYSTEM Drops were deposited on solid substrates using a pipette (Thermo Scientific, Finnpipette F2). In order to probe a velocity regime that spans 5 orders of magnitudes, we imple- mented 3 different motorized stages to drive the relative motion between the drop and the substrate. The low-velocity regime of 2 µm s-1 to 200 µm s-1 was realized by a slow linear stage consisting of a custom-made step motor whose single steps are approximately 34.2 nm. The medium-velocity regime of 0.2 µm s-1 to 1.6 mm s-1 was achieved by using a circular disk driven by a linear motor (Faulhaber 2232 A 012 SR with transmission gear 22 EK 23014:1). The circular disk rotated at a constant angular velocity in the regime of 0.02 rpm to 0.15 rpm. The liquid drops were placed at a radius of 100 mm to the rotation center. The high-velocity regime of 6.8 mm s-1 to 33.9 mm s-1 was realized again by a linear stage consisting of a custom-made step motor, which single steps were approximately 5 µm. 64 HOW DROPS START SLIDING OVER SOLID SURFACES DROP IMAGING A high-speed camera (FASTCAM Mini UX100 type 800K-M-8G, Photron) equipped with a manual zoom lens was placed orthogonal to the drop motion. Its lens was focused on the drop with the narrow side of the capillary facing this orthogonal camera to monitor the change in front and rear contact angles as well as contact length. A second camera (Marlin F131B, Allied Vision) equipped with a manual zoom lens was placed parallel to the drop motion. Its lens was focused on the drop with the wide side of the capillary facing the parallel camera to monitor the change in contact width. Using the cameras, contact angles above approximately 160° are hardly accessible due to the narrow gap between the liquid and the solid. Therefore, the values do not reflect the real contact angles but should be considered as a lower boundary. On superhydrophobic surfaces, the advancing contact angle should always be close to 180°.[61] Before the sample was moved, the glass capillary was placed in the center of the drop. Both cameras were synchronized in time with the data acquisition system. Throughout an experiment, the synchronization between force and optical measurements via the cameras was <60 ms. The drop was illuminated with two cold-light sources (KL 2500 LCD, Zeiss; L2, Leica), each of them located opposite to a camera, respectively. The contact angles, as well as the contact widths and lengths, were measured by means of ImageJ. The contact widths and lengths were calculated using the known dimensions of the capillary as scale. Figure S5–S10 show representative lateral ad- hesion forces with measured contact widths, contact lengths, and contact angles corre- sponding to the cases presented in Figure 27a. Movies S1–S7 show the motions of drops on different surfaces during the lateral adhesion force measurements corresponding to the cases presented in Figure 26 and Figure 27. AUTHOR CONTRIBUTIONS N. Gao carried out the experiments and wrote the manuscript. D. W. Pilat, N. Gao, R. Ber- ger, and H.-J. Butt designed and constructed the homebuilt setup. F. Geyer and S. Wooh prepared the solid surfaces. R. Berger, D. Vollmer, N. Gao, and H.-J. Butt contributed to experimental planning, data analysis, and manuscript preparation. All authors reviewed and approved the manuscript. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 65 ACKNOWLEDGMENTS We thank for stimulating discussions, and for technical support. This work was supported by the Collaborative Research Center 1194 (H.-J.B.), ERC advanced grant 340391 SUPRO (H.-J.B.), SPP 8173 (D.V.) and the EU Marie Sklodowska-Curie grant 722497 (D.V.). N. Gao thanks the National Postdoctoral Science Foundation of China for the International Postdoctoral Fellowship, and S. Wooh thanks the Alexander von Humboldt Foundation for a postdoctoral fellowship. D.W. Pilat is grateful for funding from the German National Academic Foundation. 66 WHEN AND HOW SELF-CLEANING OF SUPERHYDROPHOBIC SURFACES WORKS 3 WHEN AND HOW SELF-CLEANING OF SUPERHYDRO- PHOBIC SURFACES WORKS Florian Geyer,1 Maria D’Acunzi,1 Azadeh Sharifi-Aghili,1 Alexander Saal,1 Nan Gao,1,2 Anke Kaltbeitzel,1 Tim-Frederik Sloot,3 Rüdiger Berger,1 Hans-Jürgen Butt,1 and Doris Vollmer1 1Max Planck Institute for Polymer Research, Ackermannweg 10, 55128 Mainz, Germany. 2Future Industries Institute, The University of South Australia, Mawson Lake Campus, 5095 South Australia, Australia. 3Evonik Resource Efficiency GmbH, Goldschmidtstraße 100, 45127 Essen, Germany Abstract: Macroscopic contamination, such as dust, is easily cleaned away by water drops rolling across a superhydrophobic surface. Despite the enormous interest in superhydro- phobicity, it is still unclear when microscopic and, in particular, nanoscopic contamina- tion (e.g., soot, aerosols) influences superhydrophobicity and how self-cleaning works in detail. Here, we show that the pore size of the superhydrophobic surface is crucial and needs to be smaller than the contaminant size. This design principle is established using hydrophilic and hydrophobic particles varying four orders of magnitude in size (diame- ters from 80 nm to 50 µm). We gain an in-depth understanding of self-cleaning by moni- toring it on the micrometer scale using confocal microscopy and quantify the forces of drops rolling over contaminated surfaces. Outdoor environmental (>200 days) and indus- trial standardized contamination experiments demonstrate that nanoscale superhydro- phobic surfaces can resist severe contamination. Keywords: self-cleaning, superhydrophobic, contamination, nanoparticles This chapter is submitted for publication, 2019. Important Supporting Figures can be found in Appendix 10.1. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 67 INTRODUCTION Surfaces exposed to the ambient atmosphere are contaminated by particulate matter over time. Particulate matter originates from nature (e.g., microorganisms and pollen) or is man-made, such as soot from combustion.[151] This makes easy cleaning or, rather, self- cleaning desirable. Self-cleaning is one of the most prominent features of superhydro- phobic surfaces. Although almost 2000 articles were published on superhydrophobicity in 2018 alone, it is still unclear how self-cleaning works in detail and which forces are involved in the process. To obtain a comprehensive understanding, it is important to vis- ualize how the self-cleaning process evolves on the micrometer scale. The resistance of superhydrophobic surfaces to particulate matter and nano-sized contamination is inevi- table for the employment of these surfaces for a variety of real-world applications. Apart from self-cleaning,[34] for instance, this is essential for the development of water-proof textiles,[79, 139] anti-biofouling surfaces,[148, 238] drag reduction,[31, 102] anti-icing,[143, 239] mem- branes for gas exchange,[84, 116] oil/water separation,[78-79] membrane distillation,[149-150] or fog harvesting.[240-241] Thus, the need for clear design principles on how to construct sur- faces to resist particle contamination should be addressed (Figure 30). On superhydrophobic surfaces, water drops assume an almost spherical shape (contact angle, 𝜃𝜃 > 150°) and easily roll off when the surface is tilted by a few degrees (<10°).[28] To achieve superhydrophobicity, the surface has to be coated by hydrophobic nano- or mi- crometer-sized protrusions. The protrusions need to be such that air is entrapped under- neath, leading to the so-called Cassie state.[27] This results in low adhesion of drops to the surface. Macroscopic contamination can easily be removed by rinsing with water drops, e.g., during rain or artificial showers.[145] Previous contamination studies focused on the influence of millimeter- to micrometer- sized contaminants on the performance of superhydrophobic surfaces, such as hydro- phobic powder,[145] micrometer-sized particulates,[153-155] and aggregated bioparticles.[156] Here, we observe the self-cleaning process on the micrometer scale using laser scanning confocal microscopy (LSCM) and quantify the forces involved. We investigate the influ- ence of particles of various sizes (four orders of magnitude) and polarities. To design su- perhydrophobic surfaces that are resistant to hydrophilic as well as hydrophobic particle contamination, we found that the pore size of the superhydrophobic surface needs to be smaller than the contamination particle size. Even smaller nanoscopic contaminants do 68 WHEN AND HOW SELF-CLEANING OF SUPERHYDROPHOBIC SURFACES WORKS not necessarily degrade superhydrophobicity. Finally, we demonstrate that superhydro- phobic surfaces having a nanoscale pore size are capable of resisting long-term real-world exposure and industrial contamination tests. Figure 30. Illustration of contamination and self-cleaning of superhydrophobic surfaces. a) The surfaces are contaminated with particles of different sizes (80 nm to 50 µm) and polarities (hydrophobic/hydro- philic). b) Water drops roll over the contaminated Surface. c) Can the water drops remove the contamina- tion and how is superhydrophobicity affected? How does the self-cleaning evolve on the micrometer scale and which forces are involved in the self-cleaning process? RESULTS AND DISCUSSION As an example of a superhydrophobic nanoporous surface, glass slides were coated with silicone nanofilaments.[112, 115, 118, 132] Briefly, the glass slides were immersed in a mixture of toluene with a certain water content and trichloromethylsilane (TCMS). After a reaction time of 6 h, the glass slides were covered by a 1–2 µm thick layer of nanofilaments with diameters between 20 and 50 nm and spacings between 50 nm and 500 nm. To reduce the surface energy, the coated slides were fluorinated using a perfluoro silane. These surfaces show a high static water contact angle of 161° ± 4° and low roll-off angles of ≈1° for 6 µL water drops. Spherical, fluorescently labeled, hydrophilic and hydrophobic particles were used to mimic different types of atmospheric particle contaminants. Our particles covered a size range from 80 nm to 50 µm and were made of silica nanoparticles of avg. diameters 2𝑅𝑅 = 84 ± 9 nm (denoted as 80 nm particles), 210 ± 30 nm (200 nm particles), and 580 ± 120 nm (600 nm particles), silica-coated polystyrene particles of 2𝑅𝑅 = 1.45 ± 0.14 µm (1.5 µm particles), and silica-coated glass beads of 32 ± 7 µm (10–50 µm particles). Scan- ning electron microscopy (SEM) images of these particles are provided in the Supporting Information (Figure S1). Without further treatment, the particles were hydrophilic and dispersed in ethanol. To vary inherent wetting properties, a portion of the previously mentioned particles were hydrophobized using an octyl silane. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 69 Figure 31. Effect of differently sized hydrophobic particle powder contamination on nanoporous super- hydrophobic surfaces. a) Schematic illustration of the self-cleaning process of hydrophobic particle pow- ders (purple) by a water drop (gray) on a superhydrophobic surface (blue). Colors and texture were chosen to match the laser scanning confocal microscopy (LSCM) images. b) Photograph of a 10 µL water drop cleaning a nanoporous superhydrophobic surface contaminated with Oil Red dye particles (appear black). c–e) LSCM images after contaminating the nanoporous superhydrophobic surfaces (left) with powders of hydrophobic particles with diameters of 10–50 µm, 200 nm, and 80 nm (see Experimental Section 3.3.11 and Figure 36 for details of image processing). Efficient cleaning of all hydrophobic powders was verified by LSCM (center) and scanning electron microscopy (SEM, right). Scale bars: 200 nm (SEM). f,g) Contact and roll-off angles using 6 µL water drops after self-cleaning of a nanoporous surface consecutively con- taminated with hydrophobic particle powders. In order to mimic common hydrophobic contamination sources, such as soot or dust, we applied hydrophobic particle powder to the nanoporous superhydrophobic surfaces (Figure 31a,b). After contaminating the surfaces with the particle powder, the surfaces were cleaned by rinsing with 20–50 water drops of 16 ± 1 µL. These drops rolled off the inclined surfaces and took away the particles. We found that hydrophobic powder con- taminants (purple) could be effectively removed from the superhydrophobic surfaces (blue) as verified by LSCM (Figure 31c–e, Figure S2 in Appendix 10.1). SEM analysis con- 70 WHEN AND HOW SELF-CLEANING OF SUPERHYDROPHOBIC SURFACES WORKS firmed the results. Hardly any contamination could be found within or on top of the coat- ing for all particle sizes (Figure 31c–e, Figure S3 for more details of the sample contami- nated with 80 nm particles). The excellent self-cleaning is corroborated by low roll-off (<2°) and high contact angles (>150°; Figure 31f,g). Figure 32. Effect of hydrophilic particle contamination having various particle sizes deposited from eth- anol dispersion on nanoporous superhydrophobic surfaces. a) Schematic illustration of the self-cleaning process of hydrophilic particles (purple; 2R > p) deposited from ethanol dispersion by a water drop (gray). b) Particles of smaller diameter than the pore diameter (2R < p) can penetrate the coating (blue), affecting wetting properties. c–e) LSCM images (left) after contamination of the superhydrophobic surfaces with hydrophilic particles with diameters of 10–50 µm, 600 nm, and 80 nm (see Experimental Section 3.3.11 and Figure 36 for details of image processing). LSCM (center) and SEM images (right) show the surfaces after rinsing. Scale bars: 200 nm (SEM). f,g) Contact and roll-off angles using 6 µL water drops after self- cleaning of nanoporous surfaces contaminated with various hydrophilic particles (dried from ethanol dis- persion). Even after five consecutive contamination steps by particles of all different sizes followed by subsequent self-cleaning, the roll-off angles remained at 1° ± 1° and the static contact angle at 161° ± 4°. Thus, even nanoscopic hydrophobic particle powders did not affect the superhydrophobicity of the nanofilament-coated surfaces. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 71 Particulates such as water-soluble organic carbon (WSOC) are hydrophilic, can disperse in water, and can thus stick much better to surfaces.[151, 242-243] To investigate the most se- vere case of contamination, we applied hydrophilic particles dispersed in ethanol to the nanoporous surfaces (Figure 32). Fifty microliters of the particle dispersions with a con- centration of ≈10 g L-1 were cast onto the surfaces by a pipette. The dispersion wetted the surface and finally completely dried out on the surface. Here, three cases of particle sizes can be distinguished: i) The particle diameter 2𝑅𝑅 is larger than the pore diameter 𝑝𝑝 (with 𝑝𝑝 < 2𝑅𝑅), and the par- ticles cannot enter the coating (Figure 32a). The 10–50 µm and 1.5 µm particles (Figure 32c, Figure S4 in Appendix 10.1) fall in this category. In this case, the first water drop was able to remove most of the contamination. After the contaminants were re- moved, the contact angles stayed above 150°, and the roll-off angles remained below 2° (Figure 32f,g). ii) The particle diameter is similar in size to the largest pores (𝑝𝑝 ≈ 2𝑅𝑅). The 600 nm parti- cles were slightly above the largest pore diameter of the nanoporous coating but in a sim- ilar size range. Even in this case, for thin contamination layers (1–2 loose particles layers), most particles could be removed. We only found some rarely distributed particles and particle aggregates using LSCM and SEM (Figure 32d, Figure S5 in Appendix 10.1). These remaining particles did not change the wetting properties (Figure 32f,g). Only in the case of a thick aggregated contamination layer, the self-cleaning of the surface failed. In par- ticular, areas of up to 10 µm thick contamination were found on the edge of the contami- nation spot because of the coffee stain effect during drying.[244] On the coffee stain ring of the dried sample, the droplets stuck to the surface, and the roll-off angle increased to 90°. iii) The particle diameter is smaller than the pore size (𝑝𝑝 > 2𝑅𝑅), and the particles can enter the coating (Figure 32b). Here, the 80 and 200 nm particles could enter and dry out within the pores (Figure 32e, Figure S7 in Appendix 10.1 for details). Even after rinsing, strong fluorescence of the nanoparticles could be observed within the coating (Figure 32e, Figure S4 in Appendix 10.1). Indeed, hydrophilic 80 and 200 nm particles led to the failure of superhydrophobicity. The contact angles decreased to ≈140° and ≈40° for the 200 and 80 nm particles, respectively, and the roll-off angles increased above 90° for both (Figure 32f,g). In contrast to the 80 and 200 nm hydrophobic particle powders, which were littered over the surface, the hydrophilic counterpart particles were applied from a dispersion. To ex- amine whether the way how we applied the particles to the surfaces caused the superior 72 WHEN AND HOW SELF-CLEANING OF SUPERHYDROPHOBIC SURFACES WORKS performance of the hydrophobic particles, we performed an additional experiment. We applied the 80 and 200 nm hydrophobic particles from dispersion in n-hexane. Now, the roll-off angles of 6 µL water drops slightly increased to 8° ± 6° and 20° ± 15° for the 200 and 80 nm particles, respectively. The large scatter of the roll-off angles indicates large local differences in particle contaminations. Our experiments reveal that nanoporous superhydrophobic surfaces can withstand most scenarios of particle contamination. Only hydrophilic and, to a certain extent, hydropho- bic particles smaller than the pore size that are dispersed in a liquid capable of wetting the surface could lead to a failure of self-cleaning. Therefore, to design superhydrophobic surfaces that are resistant to hydrophilic and hydrophobic particle contamination, the pore size needs to be below the lowest expected particle size (𝑝𝑝 < 2𝑅𝑅). Figure 33. Contamination and self-cleaning of superhydrophobic microstructured SU-8 pillars (rectan- gular, 10 µm height with 5 × 5 µm2 top areas; center-center distance of pillars: 20 µm). a) LSCM (top) and SEM (bottom) images showing a surface contaminated with 1.5 µm particles. On the left side of the SEM image, the micropillar array is only partially filled with particles (hc < hP), whereas on the right, the particles completely covered the microstructure (hc > hP). Scale bar: 10 µm (SEM). b) Roll-off angles of 6 µL water drops after contamination of the micropillar array with hydrophilic and hydrophobic particles of various sizes and subsequent self-cleaning. Many superhydrophobic surfaces rely on microstructures.[37] As a model, we investigated microstructured superhydrophobic surfaces that were made from SU-8 micropillars (rec- tangular, 10 µm height with 5 × 5 µm2 top areas; center-center distance of pillars: 20 µm) on glass slides using photolithography. The SU-8 micropillar surfaces were fluorinated FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 73 and showed a static contact angle of 154° ± 4° and roll-off angles of 9–10° for 6 µL water drops. The superhydrophobic microstructured surfaces were exposed to hydrophilic 10– 50 µm and 1.5 µm particles dispersed in ethanol (Figure 33 and Figure S10a–c). We found that water drops could easily remove the 10–50 µm particles. The 1.5 µm particles were stuck between the micropillars since 𝑝𝑝 > 2𝑅𝑅 (Figure 33a). However, whether this causes degradation of the surface depends on the filling height of the contaminants ℎc compared to the pillar’s height ℎP. If the particles only partially fill up the space between the pillars (ℎc < ℎP), the superhydrophobicity will remain intact (Figure 33b) because the drop only rests on the top faces of the pillars (Figure S10d,e). If the coating is filled up or covered with particles (ℎc ≥ ℎP), superhydrophobicity will break down. This emphasizes the need for a small pore size to prevent small particles from penetrating the coating. Interestingly, exposing a surface to hydrophobic particle powders led to decreasing roll-off angles with each contamination step (Figure 33b). The reason is that residual particles of the powders remained on the surface, especially on the pillars’ top faces (Figure S11 in Appendix 10.1). These particles add a second scale of roughness to the surface and thus yield a hierar- chical nano/microstructured surface. Therefore, nanoscopic hydrophobic contamination can even reduce the roll-off angles of the coating. However, while the ability to withstand various kinds of contamination is essential, the ease of self-cleaning is equally crucial. How are individual particles removed from the surfaces by single drops, and how large are the involved forces? To image the pickup pro- cess of particles, we modified a confocal microscope with a stage holding a needle. This setup allowed us to move drops over the surfaces during imaging. We investigated the self-cleaning process of hydrophilic 10–50 µm particles deposited on a nanoporous super- hydrophobic surface using 10 µL drops (Figure 34a and Movie S1; particle fluorescence: purple; drop fluorescence: navy blue; reflections: light blue). Confocal microscopy ena- bled us to measure the local contact angle on the surface of the micrometer-sized particles (Figure 34b,c). When the drop was moved over the contaminated surface, individual par- ticles were picked up at the advancing contact line (Figure 34a, top). Upon further move- ment of the drop, these particles accumulated at the lower air-water interface of the drop (Figure 34a, center). The drop rested on these hydrophilic particles. The air-water-parti- cle contact lines were pinned at the lower side of the hydrophilic particles and showed contact angles of 𝜃𝜃 ≈ 40°–50° (Figure 34b). At the receding side, the drop lifted the parti- cles from the surface, yielding a layer of particles on the droplet, and a so-called liquid 74 WHEN AND HOW SELF-CLEANING OF SUPERHYDROPHOBIC SURFACES WORKS marble formed (Figure 34a, bottom).[185, 197] The same behavior was observed for hydro- phobic 10–50 µm particles (Figure 34c and Movie S2). However, the hydrophobic particles exhibited contact angles of 𝜃𝜃 ≈ 90°–100°, and the water drop rested on the top half of the particles. The experiments with both hydrophilic and hydrophobic particles clearly showed an air gap underneath the drop. The gap distance 𝐻𝐻 can be calculated using 𝐻𝐻 = 𝑅𝑅(1 − 𝑐𝑐𝑐𝑐𝑐𝑐𝜃𝜃). For a sufficient number of large particles of 𝑅𝑅 ≈ 20 µm and a contact angle of 100°, we calculated 𝐻𝐻 ≈ 24 µm. This value is in good agreement with the confocal image (Figure 34c). Smaller-sized particles were picked up similarly; however, the gap cannot be resolved by LSCM (Movie S3). To remove particles from a superhydrophobic surface, the capillary force 𝐹𝐹C acting on the particle at the rear side of the drop needs to overcome the adhesion force between the particle and the surface, 𝐹𝐹Adh. When the drop recedes and starts pulling on a particle that is adhering to the substrate at the bottom and to the air-water interface at the top side, the interface is deforming and a meniscus forms (Figure 34d). The meniscus acts as a spring with spring constant 𝑘𝑘[230, 245-246] 2𝜋𝜋𝛾𝛾 𝑘𝑘 ≈ (24) 0.81 − ln(𝑅𝑅⁄𝑙𝑙C) and pulls on the particle with a capillary force 𝐹𝐹C = 𝑘𝑘δ. Here, δ is the deformation of the surface of the drop in normal direction, 𝛾𝛾 is the surface tension of water, 𝑅𝑅 is the radius of the particle, and 𝑙𝑙C = �𝛾𝛾⁄𝑚𝑚𝑔𝑔 = 2.7 mm is the capillary length of water. The capillary force pulls on the particle until the adhesion force between the substrate and the particle is overcome, that is for 𝐹𝐹𝐶𝐶 ≥ 𝐹𝐹Adh. The adhesion force between a flat surface and a spher- ical particle can be estimated by the Johnson-Kendall-Roberts (JKR) theory:[247] 𝐹𝐹 3Adh ≈ 𝜋𝜋𝑅𝑅𝜋𝜋∆𝛾𝛾 (25) 2 Here, ∆𝛾𝛾 is the work of adhesion between the hydrophobic substrate and the particle sur- face (see Experimental Section 3.3.15). We estimate it to be in the order of ∆𝛾𝛾 = 0.11 N m-1. To take into account that superhydrophobic surfaces are rough, and roughness greatly reduces the effective contact area between the particle and the surface, we added ϕ, which is the surface fraction in contact with the particle surface as compared with the apparent contact area. From SEM images, we estimate it to be ϕ ≈ 0.2. The work to pick up a single particle 𝑊𝑊s is given by the deformation δ of the air-water meniscus before the adhesion force is overcome (Figure 34d): FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 75 𝐹𝐹2 𝐹𝐹2 𝑊𝑊s = 1𝑘𝑘𝛿𝛿2 = 1𝑘𝑘 Adh = Adh2 (26) 2 2 𝑘𝑘 2𝑘𝑘 When the drops move a distance ∆𝑥𝑥 horizontally on the superhydrophobic surface, the work required is Δ𝑊𝑊 = 𝑘𝑘𝑤𝑤𝑊𝑊sΔ𝑥𝑥, accordingly. Here, 𝑤𝑤 is the number of particles per unit area, and 𝑘𝑘 is the width of the apparent contact area of the drop. Now we can estimate the amount of the lateral adhesion force from: Δ𝑊𝑊 𝑘𝑘𝑤𝑤𝐹𝐹2 𝐹𝐹 = = 𝑘𝑘𝑤𝑤𝑊𝑊 = AdhL s (27) Δx 2𝑘𝑘 We estimate the particle density to be approximately 0.1 R-2. Thus we can estimate the lateral force caused by the removal of particles from a surface: 1 1 2 𝑘𝑘𝜋𝜋2Δ𝛾𝛾2 𝐹𝐹L ≈ 𝑘𝑘 � 3 2 𝜋𝜋𝑅𝑅𝜋𝜋∆𝛾𝛾� = 1.1 (28) 10𝑅𝑅 2𝑘𝑘 2 𝑘𝑘 Here, we further neglected lateral interactions between particles and assumed that each particle is detached independently from its neighbor. With 𝑘𝑘 ≈ 0.5 mm and 𝑘𝑘 = 0.082 N m-1 for 𝑅𝑅 = 25 µm, the lateral adhesion force is 𝐹𝐹L ≈ 3 µN (see Experimental Section 3.3.15 for a more detailed discussion). This force can be compared to the force measured by our custom-built droplet adhesion force instrument (DAFI).[77] Here, a capil- lary drags a liquid drop over a surface (Figure 34e), and the lateral adhesion force is meas- ured using a force sensor (see Chapter 2 and Experimental Section 3.3 for details). The force required to pick up a thin contamination layer (<50 µm) of hydrophilic particles us- ing a 10 µL water drop ranged from 2 µN to 5 µN (Figure 34f). Thus, the estimated values agree well with the experimental results. This agreement indicates that the model cap- tures the essential process. Note that the lateral friction force only slightly exceeded the friction force on the pristine substrate (≈1 µN). The static lateral adhesion force is over- come when a drop starts rolling off a surface: 𝐹𝐹L = 𝐹𝐹g = 𝑚𝑚𝑚𝑚 𝑐𝑐𝑖𝑖𝑤𝑤𝛼𝛼. Here, 𝑚𝑚 = 10 mg is the mass of the water drop and 𝛼𝛼 is the roll-off angle. The friction force on slightly contami- nated surfaces remained below 5 µN, corresponding to a roll-off angle ≤ 3°. These low roll-off angles demonstrate that the surfaces can be easily cleaned. The drops roll off the contaminated surface and simultaneously clean it. 76 WHEN AND HOW SELF-CLEANING OF SUPERHYDROPHOBIC SURFACES WORKS Figure 34. Illustration of the self-cleaning of a contaminated superhydrophobic surface using confocal microscopy and friction force measurements. a) A 10 µL water drop (dyed with ATTO 488, navy blue) is dragged over a nanoporous superhydrophobic surface contaminated with 10–50 µm hydrophilic particles (purple). The interface between the drop and the surface is monitored by LSCM. Particle contamination is completely taken along by the water drop (see Experimental Section 3.3.11 and Figure 36 for details of image processing). b,c) High magnification LSCM images showing the contact angle θ of the hydrophilic and hydrophobic particles in contact with water. Smaller particles lost contact with the solid surface. d) Sketch of the pickup process of particles. The deformed meniscus pulls on the particle. e) Macroscopic observation of a 10 µL drop being dragged over a surface heavily contaminated with hydrophilic 10–50 µm particles. f) Force required to clean a surface contaminated with hydrophilic 1.5 µm and 600 nm particles. The drop is moved at a velocity of v = 250 µm s-1. g) Effect of the thickness of the contamination layer (<0.1 mm, 0.1–0.2 mm, and 0.2–0.4 mm) for hydrophilic 10–50 µm particles on the force required to clean the surface. For strongly contaminated surfaces (0.2–0.4 mm contamination layer), a continuous increase in the force during the self-cleaning can be observed (1 and 2). Upon complete coverage of the drop’s surface with particles (between 3 and 4), a sudden increase in force can be observed. Drop velocity: v = 250 µm s-1. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 77 Whereas a thin layer of hydrophilic 10–50 µm particles (1–2 layers of particles, Figure 34g, thin) can easily be removed from the surface, the force increased with an increasing con- tamination layer thickness (≈3–6 layers of particles, Figure 34g, medium). For a thick con- tamination layer (≈8–12 layers), the force increased sharply to 100 µN at 𝑡𝑡 = 50 s (Figure 34g, thick). This increase corresponds to a travel distance of 𝑐𝑐 = 𝑣𝑣 𝑡𝑡 = 1.3 cm, for a drop velocity of 𝑣𝑣 = 250 µm s-1. Within this distance, the drop performed approximately 1.5 revolutions, as 2𝜋𝜋𝑅𝑅Drop ≈ 0.8 cm. After this distance, the drop is surrounded by a layer of particles and particle aggregates, as verified by video microscopy (Figure 34e). The particles rapidly leave the contact area, trying to provide a homogeneous coverage of the surface. This pro- ceeds until no more particles can be taken up and leads to interfacial jamming. As soon as the drop is covered by a layer of particles, a liquid marble is formed. After that, the particle-loaded drop rolls over particles. This is associated with a force of 100 µN, which is 10 to 20 times higher than the lateral adhesion force for a drop only partially coated with particles (roll-off angle > 90°). Thus, thick contamination layers may not be cleaned with a single drop. Consequently, the drop size needs to be increased, or several drops are required to clean the surface. To compare the results based on model contamination to real-world outdoor exposure, we fixed several coated fabrics on a car. These so-termed superomniphobic fabrics re- pelled water and organic liquids and were chosen because of their robustness.[113, 116] They were obtained by coating polyester fabrics with the same nanostructured coating consist- ing of silicone nanofilaments. Fluorination renders the superhydrophobic fabrics super- omniphobic (Figure 35a,b). Various liquids like water, coffee, wine, and hexadecane showed high receding contact angles and typically rolled off below 10°. The fabrics on the car were exposed to the atmosphere of the Rhineland-Palatine area in Germany for 257 days within a period of 426 days, covering all seasons and exposure to high UV radia- tion, rain, frost, and icing (temperatures ranged between -10 °C and 32 °C, and humidities ranged between 40% and 100%; Table S1 summarizes the temperatures, humidities, and rainfall during the exposure periods). This also included daily driving at velocities of 60 km h-1 and maximal velocities on the highway of up to 140 km h-1. During this period, the owner drove more than 5000 km. The car was parked outside, i.e., not in a garage. The experiment started beginning of October 2016. The fabrics were fixed at the front window (A-pillar), side mirror, rear side window, and back window of the car (Figure 35c and Figure S12 in Appendix 10.1). 78 WHEN AND HOW SELF-CLEANING OF SUPERHYDROPHOBIC SURFACES WORKS Figure 35. Real-world contamination test through outdoor exposure of superomniphobic fabrics fixed on a car for 257 days. a) Photograph of 20 µL drops of water (stained with methylene blue), coffee, wine, and hexadecane on a superomniphobic fabric. b) SEM images of a coated superomniphobic polyester fab- ric at different magnifications. c) Superomniphobic fabric fixed on the side mirror of the car. The fabric remained white even after 257 days of outdoor exposure. d,e) Receding contact angles and roll-off angles of 5 µL water drops in the course of the outdoor exposure of 257 days within a period of 426 days. Periods of outdoor exposure are marked in gray. f,g) Receding contact angles and roll-off angles of 5 µL hexade- cane drops in the course of the outdoor exposure. Periods of outdoor exposure are marked in gray. h) SEM image of a superomniphobic fabric after 257 days of outdoor exposure. i) Higher magnification SEM image of a dirt particle on a superomniphobic fabric. j) High magnification SEM image of the nanofilaments on a superomniphobic fabric after 257 days of outdoor exposure. Superomniphobic fabrics with slightly different wetting properties were used to account for variances in the fabrication of the surfaces. Therefore, we observed roll-off angles for water and hexadecane ranging between 2° and 4° and between 9° and 13° for 5 µL water FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 79 and hexadecane drops, respectively (Figure 35d–g). Roll-off and receding contact angles were measured after 4, 38, 79, 139, and 257 days of outdoor exposure. In between these days, the superomniphobic fabrics were removed, and the measurements were con- ducted. Even after 257 days of exposure, the roll-off angles stayed below 25° and 50° for water and hexadecane, respectively. The superomniphobic fabric fixed at the front of the side mirror experienced the most wear and contamination under the impact of insects and dirt. SEM analysis of the fabrics taken after 426 days without any further rinsing or cleaning showed only sparse particle contamination (in the 10 µm range) (Figure 35h,i). No nanocontamination between the nanofilaments could be detected (Figure 35j). Likely, contamination was removed by morning dew or rain. For example, particles can dissolve in the water or stick to the air-water interface and roll off as soon as the dew drop can overcome the lateral adhesion force. Furthermore, the release of surface energy after co- alescence of neighboring dew drops can cause merged drops to jump off the surface,[248] taking particles along. Additionally, rain periods contribute to the self-cleaning of the sur- faces. Likely, the major part of the reduction in repellency properties was due to mechan- ical damages (Figure S13). The results of the tests mentioned above were further supplemented by an industrial dirt pickup simulation test for paints and coatings from Evonik Resource Efficiency GmbH (Figure S14 in Appendix 10.1). Here, superomniphobic fabrics and benchmarks were ex- posed to a dirt treatment inside an industrial dishwasher. The dirt simulation process and dirt mixture are adapted from a method described in European patent application EP1302765. Afterward, the surfaces were cleaned in a washing step by spraying distilled water. The whole procedure lasted 3 h, which effectively corresponds to 18 months of outdoor exposure. Finally, the change in whiteness Δ𝑘𝑘 was measured. The superomni- phobic fabrics retained their whiteness, showing only a minor change of Δ𝑘𝑘 = 8 ± 1 (Figure S14c in Appendix 10.1). In comparison, an untreated fabric showed Δ𝑘𝑘 = 59, and benchmark surfaces ranged between 30 and 40. Due to the small pore size of the nanofilament coating, the superomniphobic fabrics could withstand severe environmental and industrial contamination tests and remain mostly intact after prolonged periods of exposure. This demonstrates that well-designed superhydrophobic and superomniphobic surfaces can resist real contamination scenar- ios. 80 WHEN AND HOW SELF-CLEANING OF SUPERHYDROPHOBIC SURFACES WORKS EXPERIMENTAL SECTION MATERIALS Experiments were carried out at room temperature. The following chemicals were used: trichloromethylsilane (TCMS, 99%, Sigma-Aldrich), 1H,1H,2H,2H-perfluorodecyl- trichlorosilane (PFDTS, 96%, Alfa Aesar), tetraethoxysilane (TEOS, 98%, Sigma-Aldrich), octyltrichlorosilane (97%, Sigma-Aldrich), (3-amino-propyl)triethoxysilane (98%, Sigma- Aldrich), iron(II) sulfate heptahydrate (>99%, Sigma-Aldrich), styrene (99%, Acros Organics, stabilized), acrylic acid (99.5%, Acros Organics), ammonium persulfate (98%, Acros Organics), sodium chloride (99.8%, Riedel-de Haёn), sodium hydroxide (98%, Sigma-Aldrich), ammonia (25%, VWR Chemicals), hydrogen peroxide (35 wt%, VWR Chemicals), SU-8 3010 photoresist (MicroChem), mr-Dev 600 developer (micro resist technology), n-hexane (99.99%, Fisher Chemical), toluene (99.99%, Fisher Chemical), m-xylene (98%, Sigma-Aldrich), isopropanol (99.5%, Fisher Chemical), ethanol (absolute, 99.96%, VWR Chemicals), hexadecane (99%, Sigma-Aldrich), methylene blue (>82%, Sigma-Aldrich), rhodamine B isothiocyanate (mixed isomers, Sigma-Aldrich), Nile red (Sigma-Aldrich), and ATTO488 (ATTO-TEC). Reagents were used as received. Wine and coffee were purchased from the local supermarket. Polyester fabrics (76 g m-2, weave “Crêpe de chine”) were obtained from Karstadt. Glass beads (10–50µm) were obtained from Duke Scientific Corp. Glass slides of 24 × 60 mm2 and a thickness of 150 ± 5 µm were obtained from Menzel-Gläser. Purified Milli-Q water from Sartorius Arium Pro VF was used for all experiments. PREPARATION OF THE SUPERHYDROPHOBIC NANOPOROUS GLASS SLIDES The nanoporous superhydrophobic glass slides were prepared as previously reported.[116, 118, 233] Typically, glass slides were plasma-cleaned and activated in an oxygen plasma chamber (Diener Electronic Femto, 120 W, 5 min, 6 cm3 min-1 oxygen flow rate). TCMS was mixed with toluene having a water content of 180 ± 10 ppm (400 µL TCMS per 100 mL toluene). The water content was evaluated using a Karl Fischer coulometer (Mettler Toledo C20 Compact KF coulometer). The solution was stirred for 60 s. Afterward, the activated glass slides were immersed in the solution, and the reaction chamber was sealed. After 6 h, the TCMS-coated glass slides were rinsed with n-hexane and dried under a nitrogen stream. To modify the TCMS-coated glass slides with PFDTS, they were activated in an oxygen plasma chamber (Diener Electronic Femto, 120 W, 2 min, FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 81 7 cm3 min-1 oxygen flow rate). Subsequently, 50 µL of PFDTS was mixed with 100 mL of n-hexane. The activated substrates were immersed in the solution for 20 min, rinsed with n-hexane, and dried under a nitrogen stream. Details on the synthesis of silicone nanofilaments can be found elsewhere.[112, 115, 120, 127] PREPARATION OF THE SUPEROMNIPHOBIC FABRICS The superomniphobic fabrics were prepared as previously reported (see also Chapter 4).[116] Polyester fabrics were cleaned by ultrasonication in ethanol. The dried pol- yester fabrics were coated with nanofilaments by adding 1400 µL of TCMS to a reaction chamber containing 350 mL of toluene with a water content of 150–160 ppm. The water content was evaluated using a Karl Fischer coulometer (Mettler Toledo C20 Compact KF coulometer). The solution was stirred for 60 s. Afterward, fabrics were immersed in the solution, and the reaction chamber was sealed. After 3 h, the TCMS-coated fabrics were rinsed with n-hexane and dried under a nitrogen stream. To modify the TCMS-coated fab- rics with PFDTS, they were activated in an oxygen plasma chamber (Diener Electronic Femto, 120 W, 2 min, 7 cm3 min-1 oxygen flow rate). Subsequently, 180 µL of PFDTS was mixed with 350 mL of n-hexane. The activated substrates were immersed in the solution for 20 min, rinsed with n-hexane, and dried under a nitrogen stream. PREPARATION OF THE SUPERHYDROPHOBIC SU-8 MICROPILLARS SU-8 micropillar arrays were prepared on glass slides by photolithography as previously reported.[62, 77] The rectangular pillars were designed 10 µm high with 5 × 5 µm2 top areas. The pillar–pillar distance between the centers of two adjacent pillars in a row was 20 µm. Details are described in Experimental Section 2.3.4. The SU-8 micropillars were activated in an oxygen plasma chamber (Diener Electronic Femto, 30 W, 2 min, 6 cm3 min-1 oxygen flow rate). Subsequently, PFDTS was mixed with n-hexane (50 µL PFDTS per 100 mL n-hexane). The activated SU-8 micropillar-coated glass slides were immersed in the solu- tion, and the reaction chamber was sealed. After 20 min, the fluorinated pillars were rinsed with n-hexane and dried under a nitrogen stream. SYNTHESIS OF A FLUORESCENT SILICA SHELL ON 10–50 µM GLASS PARTICLES M. D'Acunzi coated the particles according to the procedure described in her disserta- tion,[249] which is based on the procedure reported by Verhaegh et al.[250] Glass beads (2 g; 10–50 µm) were washed and activated using the Fenton reagents.[118, 251-252] The beads were dispersed in a solution prepared by slowly dissolving (caution!) iron sulfate heptahydrate 82 WHEN AND HOW SELF-CLEANING OF SUPERHYDROPHOBIC SURFACES WORKS (2 g) in 35 wt% hydrogen peroxide solution (20 mL). After 1 h, the particles were washed 3 times with water and 3 times with ethanol. Finally, they were dispersed in ethanol (80 mL), and reagents were added in this order: ammonia (6.4 mL), APS-RITC (0.17 mL; rhodamine B isothiocyanate coupled to (3-aminopropyl)triethoxysilane, see ref.[249-250])), and TEOS (0.480 mL). The mixture was kept under stirring for 24 h and then washed 3 times with ethanol. The particle size was analyzed by scanning electron microscopy (SEM). SYNTHESIS OF FLUORESCENT SILICA PARTICLES Fluorescent silica particles with diameters of 80 nm and 200 nm were synthesized by the same experimental procedure (Experimental Section 3.3.5). The size was tuned by varying the amount of ammonia in the reaction mixture. Ethanol (160 mL) and ammonia (9.6 mL for 80 nm particles and 12.8 mL for 200 nm particles) were homogeneously mixed, and APS-RITC (0.34 mL) was added. After 1 min under stirring, TEOS (0.96 mL) was supplied. The mixture was kept under stirring for 24 h at 500 rpm. The particle suspension was washed 4 times with ethanol. The particle size was analyzed by SEM. SYNTHESIS OF 1.5 µM POLYSTYRENE-SILICA CORE-SHELL PARTICLES WITH POLYSTY- RENE CORE DYED WITH NILE RED M. D’Acunzi prepared the polystyrene particles according to the procedure described in her dissertation,[249] which is based on the procedure reported by Zhang et al.[253]: “Polysty- rene cores (PS) were synthesized by soap-free emulsion polymerization. Before starting the synthesis, the inhibitor was removed from the styrene by washing with 3 aliquots of 2 M sodium hydroxide solution and 3 aliquots of water. The washed styrene was distilled under reduced pressure before use. A 2000 mL three-necked flask was equipped with a condenser, a PTFE stirrer, and a gas inlet. Water (1000 mL) was put in the reactor, and nitrogen was bubbled for 20 min. Ammonium persulfate (0.37 g) and sodium chloride (0.66 g) were dissolved in 10 mL water and put into the flask. After that, styrene (50 mL) and acrylic acid (0.66 mL) were added. The system was closed, and the mixture was heated to 75 °C under stirring at 350 rpm. After 1 day, the heating was turned off, and the mixture was cooled down to room temperature; particles were cleaned by at least 6 centrifugation steps with water and ethanol. The polystyrene particles were dyed using the method described by Schaertl et al.[254] It is based on the preparation of a double-phase system formed by particles dispersed in water FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 83 and Nile Red dissolved in xylene. Dye and xylene molecules diffuse through the water medium into the polystyrene particles. A solution of Nile Red (0.6 mg) in m-xylene (0.6 mL) was added to a dispersion of PS particles (1 g) in water (22 mL) and left under stirring for 1 day. After that time, the particles were swollen with xylene and dye. Xylene was re- moved by applying vacuum for 2 days. The particles were washed several times with wa- ter and ethanol and finally dispersed in ethanol (50 mL). A silica shell was grown on the particles by adding ammonia (4.2 mL) and TEOS (1 mL) and leaving the system under stirring for 1 day.” The particle size was analyzed by SEM. HYDROPHOBIZATION OF THE PARTICLES The hydrophilic particles were dried and then suspended in 50 mL n-hexane. Afterward, 50 µL of octyltrichlorosilane was added, the mixture was sonicated for 60 s, and stirred at 250 rpm for 45 min. Then, the hydrophobized particles were washed with n-hexane for 3 times and dispersed in n-hexane. CONTAMINATION EXPERIMENTS To contaminate the superhydrophobic surfaces with hydrophobic particle powders, ≈10 mg of the dried particle powder was distributed on an area of ≈1 cm2. To contaminate the superhydrophobic surfaces with hydrophilic particles, 50–100 µL of 10 g L-1 disper- sions of the particles in ethanol was slowly pipetted onto the surface. The dispersion was dried under ambient atmosphere at room temperature. Afterward, the contamination was cleaned by placing 20–50 water drops (16 ± 1 µL) on the inclined surface. The first few drops (<10) removed the most visible contamination. However, to ensure that all contam- ination that can be cleaned by water drops was removed, an excess number of drops was used. LASER SCANNING CONFOCAL MICROSCOPY Laser scanning confocal microscopy (LSCM) images were taken using an inverted confo- cal microscope (Leica TCS SP8 SMD, HCX PL APO 40x/0.85 CORR CS dry objective) using the Leica LAS X software. The surfaces were contaminated, as described in Experimental Section 3.3.9. LSCM images were taken before and after cleaning. For the visualization of the self-cleaning process, 10 µL water drops (dyed with ATTO 488, 1 mg L-1) were placed onto sparsely contaminated superhydrophobic nanoporous glass slides. The drops were slowly dragged over the surface using a metal needle (0.26 mm outer diameter, 31G) at- tached on a stage similar to the droplet adhesion force measurements. The self-cleaning 84 WHEN AND HOW SELF-CLEANING OF SUPERHYDROPHOBIC SURFACES WORKS process was recorded using the LSCM. The fluorescence of ATTO488 in water and the fluo- rescence of the dyed particles is shown in navy blue and pink, respectively. The reflection of light is shown in light blue. Reflection and fluorescence signals were recorded simul- taneously. PROCESSING OF LSCM IMAGES The laser scanning confocal images in Figure 34 were processed according to the proce- dure illustrated in Figure 36. For the creation of the three-dimensional (3D) LSCM images, Leica LAS X (3D viewer) software was used. Intensities of the reflection and fluorescence channels were adjusted for best clarity. In Figure 31c and Figure 32c, particle illustrations were added as described in Figure 36a,b. Figure 36. Processing of the LSCM images. a) Original image. Due to strong absorption and scattering of light, the upper part of the particle, as well as the fluorescence of the aqueous solution above, is not visible (black holes). b) Particle illustrations are added based on the positions of the reflection and fluorescence at the lower side of the particles. c) The missing fluorescence intensity above the particles, scale bars, and highlights are added to complement the image. DROPLET ADHESION FORCE INSTRUMENT EXPERIMENTS Droplet adhesion force experiments were conducted with a custom-built setup similar to that in Chapter 2.[77] A glass capillary was coupled to a force sensor (FTS-1000, FemtoTools; instead of the laser and PSD in Chapter 2). The sensor had a resolution in the range of a single micronewton and a time resolution of 1 kHz. To measure the lateral adhesion force, a drop was deposited on a substrate, and the drop was approached using the glass capil- lary. The substrate movement was realized using a linear stage with a customized stepper motor and gear set. The measurements were conducted at a speed of 250 µm s-1. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 85 SCANNING ELECTRON MICROSCOPY Scanning electron microscopy (SEM) images were taken with a Zeiss LEO 1530 Gemini SEM at gun voltages of 3 kV using the in-lens detector. To avoid charging, samples were sputtered with Pt before measurement using a BalTec MED 020 modular high vacuum coating system (with an argon pressure of 2 × 10−5 bar and a current of 30 mA, 7 nm Pt). CONTACT ANGLE MEASUREMENTS Static contact angles were measured using a DataPhysics OCA35 contact angle goniometer. Six microliter drops were deposited onto the substrate. The measurement was repeated in three different spots per substrate. The error of the static contact angle measurements was estimated to be between ±4° and ±8° depending on the substrate. For advancing and receding contact angles, 6 µL drops were deposited on the surface. Afterward, 20 µL of the liquid was added to and removed from the drop. The measurement was consecutively repeated three times at the same spot and at three different spots per substrate. The error of the advancing and receding contact angle measurements was es- timated to be ±4°. Roll-off angles were measured using the DataPhysics OCA35 contact angle goniometer. Therefore, 6 µL drops (6 µL for nanofilament-coated glass slides and 5 µL for nanofila- ment-coated fabrics) were deposited on the substrates, and the measuring plate was tilted until the drops rolled off. The roll-off angle was determined in at least 6 different spots per substrate, and mean roll-off angles and corresponding standard deviations were cal- culated. ESTIMATION OF LATERAL ADHESION FORCE OF A PARTICLE-COVERED SUBSTRATE This section describes the estimation of the force required to roll a drop over a layer of spherical particles of radius 𝑅𝑅 on a superhydrophobic surface. The apparent contact area between the drop and the surface is supposed to have a width 𝑘𝑘. The density of particles on the substrate 𝑤𝑤 (in number per unit area) is supposed to be lower than a full monolayer but large enough to support the drop and keep it from touching the substrate. We neglect forces between neighboring particles. When the drop moves a distance 𝑥𝑥 in lateral direc- tion particles attached to the air-water interface are lifted because the apparent contact angles are high (Figure 34). 86 WHEN AND HOW SELF-CLEANING OF SUPERHYDROPHOBIC SURFACES WORKS The work required to move the drop a distance 𝑥𝑥 is: 𝑊𝑊 = 𝑘𝑘 ∙ 𝑥𝑥 ∙ 𝑤𝑤 ∙ 𝑊𝑊s (29) Here, 𝑊𝑊s is the work required to detach a single particle from the substrate. The lateral force required to move the drop is: d𝑊𝑊 𝐹𝐹L = = 𝑘𝑘 ∙ 𝑤𝑤 ∙ 𝑊𝑊 (30) d𝑥𝑥 s To obtain 𝑊𝑊s, we consider an individual particle in contact with the substrate and being attached to the air-water interface. When the drop moves, the particle at some point will find itself at the receding side. From the perspective of the particle, the air-water interface mainly moves upwards. It starts pulling on the particle. When pulling on a particle stick- ing at a liquid-air interface in the normal direction, the air-water interface deforms. The force increases linearly with the distance the particle is moved out of equilibrium.[255-258] The air-water interface acts as a spring with spring constant k:[230, 245-246] 2𝜋𝜋𝛾𝛾 𝑘𝑘 = (31) 0.8091 − ln(𝑅𝑅⁄𝑙𝑙C) Here, 𝑙𝑙C = �𝛾𝛾⁄𝑚𝑚𝑔𝑔 is the capillary length. For water, it is 2.7 mm. Since the radius of cur- vature of the drop is much larger than the particle size, the interface can be described by a spring with spring constant 𝑘𝑘. With 𝑅𝑅 = 25 µm we get 𝑘𝑘 = 0.0082 N m-1. The capillary force pulls on the particle until the adhesion force between substrate and particle is overcome. The adhesion force 𝐹𝐹Adh between a flat surface and a spherical par- ticle is given, e.g., by the JKR theory:[247] 3 𝐹𝐹Adh = 𝜋𝜋𝑅𝑅𝜋𝜋(𝛾𝛾S + 𝛾𝛾P − 𝛾𝛾SP) (32) 2 Here, 𝛾𝛾S is the surface energy of the substrate, 𝛾𝛾P is the surface energy of the particle, and 𝛾𝛾SP is the interfacial energy between the two solids. Neglecting 𝛾𝛾SP and taking the surface energy of the fluorinated surface to be 10 mN m-1 and that of the hydrophilic oxide to be 100 mN m-1 we estimate (𝛾𝛾 -1S + 𝛾𝛾P − 𝛾𝛾SP) = 110 mN m . Since the actual contact area is re- duced on a superhydrophobic surface, we insert a factor 𝜋𝜋. 𝜋𝜋 is the ratio of the actual to the apparent contact area of the particle. For the nanofilaments used in the experiment shown in Figure 34, we estimate 𝜋𝜋 = 0.2 from SEM images. With 𝑅𝑅 = 25 µm, we get 2.6 µN. Capillary forces can usually overcome the adhesion forces. The capillary force can reach FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 87 𝐹𝐹 2 𝜃𝜃c,max = 2𝜋𝜋𝑅𝑅𝛾𝛾 cos , where 𝜃𝜃 is the contact angle the particle forms with the air-water in-2 terface. If the actually applied force exceeds this value, the liquid will fully dewet the par- ticle, and the capillary forces vanish. The air-water interface pulls on the particle until the particle detaches from the substrate. The deformation of the air-water interface at this point is 𝛿𝛿 = 𝐹𝐹Adh⁄𝑘𝑘. The work required to reach that point is: 1 𝑘𝑘𝐹𝐹2 𝐹𝐹2 𝑊𝑊s = 𝑘𝑘𝛿𝛿2 = Adh = Adh (33) 2 2𝑘𝑘2 2𝑘𝑘 To estimate 𝑤𝑤 in our experiments, we assume that the large particles dominate. Their packing will not exceed the close random packing density of 0.82⁄(𝜋𝜋𝑅𝑅2).[259] In our case, we estimate that the particle density is roughly a factor 2 lower. Thus, we take 𝑤𝑤 = 1⁄(10𝑅𝑅2) . Putting everything together for an estimate and taking a typical width of 𝑘𝑘 ≈ 0.5 mm for 10 µL-sized water drops, we obtain: 𝑘𝑘𝐹𝐹2Adh 0.0005 m ∙ (2.6 × 10−6N)2 m𝐹𝐹𝐿𝐿 = 𝑘𝑘𝑤𝑤𝑊𝑊s = 20𝑘𝑘𝑅𝑅2 = 20 ∙ 0.082 N ∙ 6.25 × 10−10 2 ≈ 3 µN m (34) AUTHOR CONTRIBUTIONS F. Geyer fabricated the surfaces, carried out the experiments and characterization, unless stated otherwise below, and wrote the manuscript. M. D’Acunzi prepared the nano- and microparticles. A. Saal, F. Geyer, and N. Gao carried out the lateral adhesion force exper- iments. A. Sharifi-Aghili fabricated and characterized the superomniphobic fabrics for the outdoor exposure experiments. A. Kaltbeitzel assisted the laser scanning confocal measurements. T.-F. Sloot was responsible for the dirt dishwasher experiments. H.-J. Butt and D. Vollmer devised the theoretical calculation of the lateral adhesion force. F. Geyer, D. Vollmer, R. Berger, and H.-J. Butt contributed to experimental planning, data analysis, and manuscript preparation. All authors reviewed and approved the manuscript. ACKNOWLEDGMENTS D. Vollmer kindly provided her car for the outdoor experiments. We thank for carefully reading the manuscript. We thank for technical support. This work was supported by the ERC Advanced Grant No. 340391 “SUPRO” (H.-J.B., D.V., F.G.), the German Research Foundation (DFG) with the Collaborative Research Center 1194 (H.-J.B., F.G.), the European Union’s Horizon 2020 research and innovation program LubISS No. 722497 (D.V.). 88 ENHANCING CARBON DIOXIDE CAPTURE USING ROBUST SUPEROMNIPHOBIC MEMBRANES 4 ENHANCING CARBON DIOXIDE CAPTURE USING RO- BUST SUPEROMNIPHOBIC MEMBRANES Florian Geyer,1 Clarissa Schönecker,1 Hans-Jürgen Butt,1 and Doris Vollmer1 1Max Planck Institute for Polymer Research, Ackermannweg 10, 55128 Mainz, Germany Abstract: Superomniphobic membranes for post-combustion CO2 capture are introduced. Concentrated aqueous amine solutions stay on the topmost part of the membranes, providing a large liquid-CO2 interface. Wetting of the membrane, which reduces the cap- ture efficiency, is prevented. The CO2 capture rates using the chemically, mechanically, and thermally stable superomniphobic membranes are enhanced by up to 40% relative to commercial membranes. Keywords: wetting, superhydrophobic, superoleophobic, CO2 capture, membrane This chapter is reprinted with permission from F. Geyer et al., Adv. Mater. 2017, 29, 1603524. Copyright 2017, John Wiley and Sons The Supporting Information is available from Wiley Online Library: https://doi.org/10.1002/adma.201603524 FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 89 INTRODUCTION The atmospheric concentration of CO2 has increased substantially in recent decades.[178, 260] This increase results from both natural and anthropogenic emissions. Major anthropo- genic sources of CO2 include cement production, the iron and steel industry, gas flaring, and combustion of fossil fuels. Slowing global warming requires a substantial reduction of CO2 emission. However, according to the recent report “Global Carbon Budget,” the world pumped 39.8 billion tons of CO2 into the air in 2014.[261] Despite urgent warnings about the necessity to curb greenhouse gases, this is a 2.3% increase in CO2 release com- pared to 2013. The search for alternative or improved methods to reduce CO2 emission and to capture and store released CO2 has stimulated intense research in academia and industry.[170, 262-264] Today so-called absorbers are used in power plants for post-combustion capture of CO2. In an absorber, an aqueous alkanolamine solution flows downward while flue gas streams upward.[177-178] The tertiary alkanolamine N-methyl diethanolamine (MDEA), often pro- moted by piperazine,[265] shows one of the best combinations of high loading capacity, low volatility, and good thermal stability. For further processing or storage, CO2 is released by heating the alkanolamine solution to above 100 °C.[266] To keep the thermal energy re- quired for releasing the captured CO2 low, the alkanolamine concentration should be as high as possible. This increases the overall CO2 capacity of the capture medium. However, the loading ratio αLoad, which is the amount of bound CO2 molecules per amine molecule, decreases at concentrations above 20 wt%. Beyond ≈50 wt%, the loading ratio becomes so low that the overall capacity decreases by adding more MDEA.[267] Therefore, alkanola- mine concentrations between 20 and 50 wt% are most commonly used depending on the specific requirements. As an alternative to absorbers, we focus on CO2 capture using gas contactor mem- branes.[171, 173] Here, a porous membrane spatially separates the gas and liquid streams. The flue gas is pumped across one side of the membrane and the capturing solution across the other. Advantages are its modular, easily-scalable nature[173-174] and compactness facil- itating process intensification.[268-271] In addition, liquid and gas flow rates can be inde- pendently tuned and controlled. Even at high flow rates, dragging of drops from the cap- turing media into the CO2 stream is reduced as the liquid and gas phase are separated by the membrane.[173-174] However, three challenges need to be addressed before this technol- ogy can be implemented on an industrial scale:[173, 176] i) Wetting of membrane pores. Wet- ting of the membrane pores greatly lowers the membrane performance as gas diffusion 90 ENHANCING CARBON DIOXIDE CAPTURE USING ROBUST SUPEROMNIPHOBIC MEMBRANES is greatly reduced in the almost stagnant liquid. ii) Long-term stability. Most membrane materials show an insufficient chemical resistance to the alkaline capturing solution, typ- ically accompanied by poor mechanical and thermal stability resulting in a breakdown of the membrane and thus a complete failure. iii) Clogging. Often, particles or other compo- nents present in the flue gas cause clogging of the pores. Thus, mass transfer is reduced accordingly. To reduce wetting of the pores, the membrane material needs to be hydrophobic.[13] The most common materials are polypropylene (PP), polyvinylidene fluoride (PVDF), and pol- ytetrafluoroethylene (PTFE).[173-174] Progress has been reported using superhydrophobic membranes, such as polypropylene,[272] fluorinated PDMS/PVDF,[273], or fluorinated ce- ramic membranes.[274] However, alkanolamine solutions have a low surface tension and wet superhydrophobic surfaces. To circumvent wetting, the gas flow can be increased,[274] causing gas bubbles to form in the capturing medium. However, this leads to foaming, transfer of particulate matter and gaseous impurities into the amine solution.[173] There- fore, the main challenge is to find membranes that repel concentrated amine solutions. These membranes must also show good chemical stability as amine solutions are strongly basic and thus highly corrosive. Highly liquid-repellent membranes — so-called superomniphobic membranes ― may be the key to overcoming the limitations of conventional gas contactor membranes. Super- omniphobicity requires that the surface is coated by overhanging hydrophobic nano- or micrometer-sized protrusions.[41, 82, 275] When placing an aqueous or non-aqueous drop on such a surface, the drop takes an almost spherical shape and rolls off when the surface is tilted by less than 10 degrees.[41, 82] On superomniphobic surfaces, the liquid does not wet the surface but stays on top of the overhangs, entrapping gas underneath. This results in a liquid-gas contact area close to 100%.[84, 87-88] The large liquid-gas contact area potentially translates into a high CO2 capture rate. Here we show the potential of superomniphobic membranes for CO2 capture. We present the first superomniphobic membranes that pre- vent wetting of membrane pores by highly concentrated alkanolamine solutions, despite being highly alkaline. Due to their superomniphobic nature, the membranes provide a large liquid-gas interface for efficient gas exchange. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 91 RESULTS AND DISCUSSION We fabricated chemically and mechanically robust superomniphobic flat-sheet mem- branes based on a polyester fabric (Figure 37 and Figure S1 and S2, in the following, de- noted as superomniphobic membrane or SOM). The 160 µm thick fabrics comprise inter- woven networks of fibers with diameters between 8 and 15 µm (Figure 37a–c). To render the hydrophilic polyester fabrics superomniphobic, we first immersed the fabrics in tol- uene containing trichloromethylsilane and trace amounts of water (≈150 ppm). Trichloro- methylsilane hydrolyzes and reacts with the hydroxy groups on the polyester surface, forming silicone nanofilaments (Figure 37d–f, Figure S3 summarizes the involved chem- ical reactions).[67, 112, 115, 120, 133] After a reaction time of 3 h, the fibers were surrounded by a 1–2 µm thick layer of silicone nanofilaments with a diameter of 20–80 nm (Figure 37f). Figure 37. a–f) Morphology of the superomniphobic mem- brane. Scanning electron mi- croscopy (SEM) images of bare (a–c) and coated (d–f) polyes- ter fabrics at different magnifi- cations. f) Higher magnifica- tion SEM image of the coated polyester shows the nanofila- ments on a polyester fiber. g,h) SEM images, taken under an angle of 45°, reveal the mi- crometer-sized pores of the su- peromniphobic membrane. The pore size ranges from less than 1 µm between single fi- bers up to 80 µm between bun- dles of fibers. To achieve superomniphobicity, the fabrics were activated in an oxygen plasma and mod- ified with 1H,1H,2H,2H-perfluorodecyltrichlorosilane to reduce the surface energy. The nanofilament coating introduces a surface roughness with overhang geometry on the pol- yester microfibers (Figure 37e,f). Nitrogen adsorption-desorption isotherms showed the appearance of meso- (2–50 nm) and macropores (>50 nm) after application of the coating 92 ENHANCING CARBON DIOXIDE CAPTURE USING ROBUST SUPEROMNIPHOBIC MEMBRANES to the polyester fabric (Figure S4). The BET (Brunauer-Emmett-Teller) surface area increased from ≈0.2 m2 g-1 for the bare fabric to 3.9 m2 g-1 for the superomniphobic membrane. In addition to the pores introduced by the nanofilament coating, the super- omniphobic membrane had micrometer-sized pores ranging from less than 1 µm be- tween single polyester microfibers up to 80 µm between bundles of microfibers (Figure 37g,h and Figure S1). Mercury intrusion porosimetry measurements confirmed the presence of micrometer-sized pores (Figure S5, average pore diameter 70 µm; pores below 0.5 µm were not considered in the calculation). The coating did not change the overall morphology of the fabrics (compare Figure 37a,b to Figure 37d,e). To investigate the difference between superhydrophobic and superomniphobic mem- branes for CO2 capture, we also coated a commercial polypropylene membrane (avg. pore size 200 nm, porosity 75%, obtained from 3M) with nanofilaments (denoted as SHPP-200). As we did not fluorinate the nanofilaments, the membranes were solely superhydropho- bic but not superomniphobic. Figure S6 shows SEM images of the superhydrophobic membrane. The wetting behavior of the membranes was tested using liquids of greatly different surface tension and polarity, including water, hexadecane, and concentrated MDEA solutions (Figure 38 and Table S2). To quantify the wetting behavior of the mem- branes, we measured receding contact angles and roll-off angles, as the static contact an- gle depends on how a drop is deposited on a surface (see Experimental Section 4.4.5 for details, contact angles and contact angle hystereses are summarized in Table S2). The re- ceding contact angle 𝜃𝜃Rec scales with the adhesion of the liquid to the substrate. The ad- hesion is high if the solution wets the membrane.[87] On the polyester fabrics coated with fluorinated silicone nanofilaments all liquids showed high receding contact angles (>150°) and low roll-off angles (<10°), demonstrating their superomniphobicity for liquids with surface tensions greater or equal to 27.5 mN m-1 (Figure 38d,e and Movie S1 and S2). Laser scanning confocal microscopy revealed that the nanofilament coating is not wetted by MDEA solutions (Figure 38c). This is proven by the extremely high contact angle (176°) and the air-cushion between the glass substrate and the MDEA (50 wt%) droplet (see Ex- perimental Section 4.4.8 for further details). On the polypropylene membranes coated with nanofilaments (SHPP-200), water showed high receding contact angles (>150°), and water drops easily rolled off (<10°), demonstrating their superhydrophobicity. However, the super-liquid-repellency broke down for MDEA solutions (roll-off angle > 40°). FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 93 Figure 38. Wetting properties of the superomniphobic, superhydrophobic, and commercial membranes. a) Photograph of 15 µL drops of water (stained with methylene blue), aqueous MDEA solutions (20 wt%, colorless and 50 wt%, phenolphthalein) and hexadecane (oil red) on a superomniphobic membrane. b) Surface tensions of the investigated liquids. c) Laser scanning confocal microscope image of an MDEA solution (50 wt% with ATTO488 dye) drop on a nanofilament-coated superomniphobic glass slide. Reflec- tions are shown in blue and fluorescence in pink. d) Receding contact angles for water, MDEA solutions (20 and 50 wt%), and hexadecane on the superomniphobic (blue), superhydrophobic (violet), and the commercial membranes (green, red, black). e) Roll-off angles for 6 µL water, MDEA solution (20 and 50 wt%), and hexadecane drops on the superomniphobic, superhydrophobic, and the commercial mem- branes. The color code is identical to d). For error bars, see Table S1 and Table S2. In contrast, state-of-the-art commercial polypropylene membranes (avg. pore size 200 nm, porosity 75%, denoted as PP-200, obtained from 3M) and polytetrafluoroethylene membranes (avg. pore size 220 nm denoted as PTFE-220; avg. pore size 1 µm denoted as PTFE-1000, both 85% porosity, obtained from Merck Millipore) were wetted by the liquids (Figure 38d,e and Movie S3). This is reflected in the low receding contact angles (<110°) and high roll-off angles (>70°) for all liquids except water. Hexadecane immediately spread on the superhydrophobic and on all the commercial membranes (Movie S4). Figures S7–S9 show SEM images of the commercial membranes. Uptake of CO2 was quantified using a custom-made attenuated total reflection Fourier transform infrared (ATR-FTIR) spectrometer (Figure 39a). The ATR-FTIR spectrometer 94 ENHANCING CARBON DIOXIDE CAPTURE USING ROBUST SUPEROMNIPHOBIC MEMBRANES had a built-in flow cell to monitor the CO2 uptake in solution directly. To measure the CO2 exchange performance, a membrane was inserted between the upper and lower parts of the exchange chamber and fixed in a holder (Figure 39b). The gas diffuses through the superomniphobic membrane, whereas the liquid flows over the topmost part of the fluorinated nanofilaments (Figure 39c). For capturing CO2, a 20 wt% MDEA solution was used. A small part of the flow was split off and ran parallel to the main stream through the IR spectrometer (see Experimental Section 4.4.7 for details). We used a 20 wt% MDEA solution as it shows a high loading ratio and a 5× lower viscosity than a 50 wt% solution. The lower viscosity reduces the flow resistance.[276] Figure 39. Setup for quantifying CO2 up- take. a) Sketch illustrating the CO2 exchange setup, including the ATR-FTIR spectrometer. b) Layout of the two halves of the exchange chamber showing inlets and outlets as well as flow channels for CO2 and liquid flow. The superomniphobic membrane is placed in between. Flow velocity of MDEA solution: 2.7 cm s-1. c) Schematic of polyester fibers coated with nanofilaments. CO2 diffuses through the superomniphobic membrane into the MDEA solution. CO2 is captured in the form of bicarbonate by MDEA, according to CO2 + (CH3)(CH2CH2OH)2N + H2O → (CH3)(CH2CH2OH)2NH+ + HCO -3 . Formation of bicarbonate bands was observed in the IR spectra, e.g., at 1302 cm-1 (HCO -3 scissoring) (Figure 40a). The peaks at 1026 cm-1, 1076 cm-1, 1360 cm-1, and 1643 cm-1 correspond to MDEA, MDEAH+, HCO -3 and OH stretching, respectively.[277] The height of the bicarbonate band (𝛼𝛼) scales with the concentration of CO2 in solution. Initially, the CO2 absorption increases almost linearly with time (Figure 40b). After 180 min, the increase in CO2 absorption slows down, indicating saturation of the solution with CO2. Please note that the time to reach saturation depends on the interplay between the volume of the capture solution and the gas ex- change area through the membrane. Here, it was chosen to enable a good time-resolution of the capturing process. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 95 Figure 40. CO2 absorption and exchange performance. a) Time-evolution of FTIR spectrum of MDEA so- lution (20 wt%) over 60 min gas exchange through a superomniphobic membrane. Colored graphs show the temporal changes of the spectrum in 12 min intervals (bottom to top). α denotes the bicarbonate band at 1302 cm-1. b) CO2 absorption over 390 min of gas exchange using a superomniphobic membrane. c) Long-term performance of the superomniphobic membrane. MDEA solution (20 wt%) was exchanged during operation after each 24 h. d) SEM images of the superomniphobic membrane after 50 h of gas ex- change revealing hardly any damage. e) CO2 absorption using the superomniphobic (blue), superhydro- phobic (violet), and state-of-the-art commercial membranes (green, red, black) over 60 min of gas ex- change. Data are each based on three independent measurements. f) Comparison of the CO2 absorption performance of the superomniphobic membrane to the superhydrophobic and commercial membranes. See Figure S10 for error bars. To investigate the long-term performance of the superomniphobic membranes, we re- placed the MDEA solution during operation by fresh solutions after 24 h and again after 48 h. The CO2 absorption rate was identical for all three runs (Figure 40c), thus even after 50 h of continuous operation, the membrane remained intact, and wetting of pores can 96 ENHANCING CARBON DIOXIDE CAPTURE USING ROBUST SUPEROMNIPHOBIC MEMBRANES be excluded (Figure 40d). The CO2 uptake via the superomniphobic membranes exceeded those of all tested state-of-the-art commercial as well as superhydrophobic membranes by up to 40% under identical conditions (Figure 40e,f). The average CO2 gas flux during the first 30 min was ≈2.5 mmol m-2 s-1 (see Experimental Section 4.4.12 for details of the calculation). Notably, the large gas transfer rate is achieved solely via diffusion of CO2 through the superomniphobic membrane. Even after 50 h of operation, we did not ob- serve any gas bubbles in the capturing solution, i.e., foaming is not a problem (Movie S5). By applying a superhydrophobic coating, the performance of the PP-200 membrane in- creased. However, in comparison to the superomniphobic membranes, the superhydro- phobic SHPP-200 membranes cannot avoid wetting by the MDEA solution. Thus the CO2 uptake via the SOM was more than 20% higher. To verify the chemical stability of the superomniphobic membranes, we exposed them to 50 wt% MDEA solutions for up to two weeks (Figure 41a and Figure S11). Even after 14 days, the membranes were not wetted. A plastron layer was still visible. With pro- longed exposure time, the receding contact angles slightly decreased. Both hexadecane and MDEA (20 wt%) solution showed roll-off angles below 30°. Such low roll-off angles document the membranes’ good chemical resistance toward amine solutions and the presence of air cushions. Exposure to the amine solution caused neither swelling nor de- formation of the membranes. The slight reduction of the receding contact angles was caused by the amine solution attacking parts of the topmost area of the coating, leading to partial detachment of nanofilaments from polyester microfibers (Figure S12). On a nan- ofilament-coated glass slide, no detachment of nanofilaments was observed after 14 days of exposure to an MDEA (50 wt%) solution (Figure S13). Thus, the polyester or the bond of the filaments to the polyester fabric was possibly hydrolyzed (Figure S14 summarizes the possible degradation mechanisms). The superomniphobic membranes kept their su- per-liquid-repellency after being heated to temperatures up to 200 °C for 24 h, demon- strating their high thermal stability (Figure 41b and Figure S15). Thus, hot flue gas streams need not be cooled to room temperature. At higher temperatures, the polyester fabric starts melting. Superomniphobic membranes have another advantage compared to hydrophobic ones. The self-cleaning nature of superomniphobicity also reduces the problem of clogging of the membrane by particles (Movie S6). Moreover, the superomni- phobic membrane showed a superior ultimate tensile strength, even though it had the lowest thickness (160 µm, see Experimental Section 4.4.11). All tensile specimens had a width of 4 mm. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 97 Figure 41. Chemical, mechanical, and thermal stability of the superomniphobic membrane. a) Receding contact an- gles and roll-off angles for 6 µL water (solid circles), MDEA (20 wt%) solution (stars), and hexadecane drops (open squares) on superomniphobic membranes that were ex- posed to MDEA (50 wt%) solutions for up to 14 days. b) Receding contact angles and roll-off angles for 6 µL drops of an MDEA (20 wt%) solution on superomniphobic membranes that were heated at the given temperatures for 24 h. c) Force-versus-strain curves. The force maxima denote the maximal strains the membranes can resist. The superomniphobic membrane was stressed in horizon- tal and vertical directions, showing the isotropy of the me- chanical strength. The rupture force was 6× higher than the supported PTFE membranes and 30× higher than the polypropylene membranes (Figure 41c). The stability of the PTFE membranes without the supporting mesh is even lower. The excellent mechanical stability of the su- peromniphobic membranes enables easy handling during operation. Even after we ap- plied a load of 25 N on a tensile specimen, the contact angles remained unaltered. Also, the stability of the nanofilaments against high liquid flow velocities was investigated by exposing the superomniphobic membrane to a stream of tap water for 60 min at an incli- nation of ≈15° relative to the stream (Movie S7). The water jet had a flow velocity of ≈220 cm s-1, which is 80× higher than the employed flow velocity for the gas exchange ex- periments. Nonetheless, after 60 min the SOM was still dry, and the receding contact an- gles and roll-off angles for an MDEA (20 wt%) solution remained identical (𝜃𝜃Rec = 158° ± 4°, 98 ENHANCING CARBON DIOXIDE CAPTURE USING ROBUST SUPEROMNIPHOBIC MEMBRANES roll-off angle = 4° ± 1°). SEM images confirmed that the nanofilaments did not detach (Figure S16). CONCLUSION In conclusion, preventing wetting of membrane pores by the capturing media is the key challenge for gas contactor membranes to achieve higher capture efficiency. Inspired by the super-liquid-repellent and self-cleaning properties of superomniphobic surfaces, one can devise superomniphobic membranes as novel gas contactor membrane material, as these can prevent wetting of the pores by the capturing solution. Simultaneously the large liquid-gas interface of the superomniphobic membranes ensures a high gas absorption flux. The fabricated superomniphobic membranes also keep their super-liquid-repel- lency for more than two weeks being exposed to 50 wt% MDEA solution and show excel- lent tensile strength. As aqueous alkanolamine solutions are strongly basic, the textured membrane has to be coated with a dense and chemically stable layer of fluorinated mol- ecules to prevent decomposition of the underlying substrate. Keeping this in mind, cheap, flexible, and mechanical robust materials such as polyester fabrics can be transformed into robust superomniphobic gas contactor membranes. Furthermore, the self-cleaning properties of superomniphobic membranes reduce fouling of the membrane by particu- late matter. Superomniphobic gas contractor membranes are not limited to CO2 capture but may also be applied to scrubbing of other gases, for example, H2S. The modular setup and compactness of gas contactor membranes render a downscaling of absorber cham- bers possible. This might also encourage capture of hazardous gases at processes releas- ing only smaller gas quantities. EXPERIMENTAL SECTION MATERIALS If not stated otherwise, experiments were carried out at room temperature. The following chemicals were used: trichloromethylsilane (TCMS, 99%, Sigma-Aldrich), 1H,1H,2H,2H-perfluorodecyltrichlorosilane (PFDTS, 96%, Alfa Aesar), n-hexane (99,99%, Fisher Chemical), toluene (99,99%, Fisher Chemical), ethanol (99.8%, Sigma-Aldrich), N-methyl diethanolamine (99%, Sigma-Aldrich), ethylene glycol (99%, Sigma-Aldrich), hexadecane (99%, Sigma-Aldrich), methylene blue (>82%, Sigma-Aldrich), oil red (>75%, FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 99 Sigma-Aldrich), phenolphthalein (Sigma-Aldrich), ATTO488 (ATTO-TEC), carbon black (99.9%, Alfa Aesar). Reagents were used as received. Polyester fabrics (polyethylene terephthalate (PET), 76 g m-2, weave “Crêpe de chine”) were obtained from Karstadt. Polypropylene (PP) membranes (ACCUREL PP 2E HF, denoted as PP-200) with an average pore size of 200 nm were obtained from 3M. Polytetrafluoroethylene (PTFE) membranes (Fluoropore FGLP and FALP) with an average pore size of 220 nm (denoted as PTFE-220) and 1 µm (denoted as PTFE-1000) were obtained from Merck Millipore. Glass slides of 24 × 60 mm2 and a thickness of 170 ± 5 µm were obtained von Menzel-Gläser. CO2-gas bottle (99.99%) was obtained from Westfalen AG. Purified Milli-Q water from Sartorius Arium Pro VF was used for all experiments. PREPARATION OF THE SUPEROMNIPHOBIC MEMBRANES Polyester fabrics were cleaned by ultrasonication in ethanol. The dried polyester fabrics were coated with nanofilaments by adding 1400 µL of TCMS to a reaction chamber con- taining 350 mL of toluene with a water content of 150 ppm. The water content was evalu- ated using a Karl Fischer coulometer (Mettler Toledo C20 Compact KF coulometer). The solution was stirred for 60 s. Afterward, fabrics were immersed in the solution, and the reaction chamber was sealed. After 3 h, the TCMS-coated fabrics were rinsed with n-hex- ane and dried under a nitrogen stream. To modify the TCMS-coated fabrics with PFDTS, they were activated in an oxygen plasma chamber (Diener Electronic Femto, 120 W, 2 min, 7 cm3 min-1 oxygen flow rate). Subsequently, 180 μL of PFDTS was mixed with 350 mL of n-hexane. The activated substrates were immersed in the solution for 20 min, rinsed with n-hexane, and dried under a nitrogen stream. Details on the synthesis of silicone nanofil- aments can be found elsewhere.[112, 115, 127] PREPARATION OF THE SUPERHYDROPHOBIC MEMBRANES Polypropylene membranes (PP-200) were activated in an oxygen plasma chamber (Diener Electronic Femto, 30 W, 1 min, 7 cm3 min-1 oxygen flow rate). The activated polypropylene membranes were coated with nanofilaments by adding 1400 μL of TCMS to a reaction chamber containing 350 mL of toluene with a water content of 180 ppm. The water con- tent was evaluated using a Karl Fischer coulometer (Mettler Toledo C20 Compact KF cou- lometer). The solution was stirred for 60 s. Afterward, the membranes were immersed in the solution, and the reaction chamber was sealed. After 3 h, the TCMS-coated mem- branes were rinsed with n-hexane and dried under a nitrogen stream. 100 ENHANCING CARBON DIOXIDE CAPTURE USING ROBUST SUPEROMNIPHOBIC MEMBRANES PREPARATION OF THE SUPEROMNIPHOBIC GLASS SLIDES Silicone nanofilaments on glass slides were synthesized by modifying previously reported methods.[112, 115] The fabrication contained the following steps:[116, 118, 233] Glass slides were plasma-cleaned and activated in an oxygen plasma chamber (Diener Electronic Femto, 120 W, 5 min, 7 cm3 min-1 oxygen flow rate). The activated glass slides were coated with nanofilaments by adding 250 µL of TCMS to a reaction chamber containing 350 mL of tol- uene with a water content of 125 ppm. The water content was evaluated using a Karl Fischer coulometer (Mettler Toledo C20 Compact KF coulometer). The solution was stirred for 60 s. Afterward, the glass slides were immersed in the solution, the reaction chamber was sealed, and stirring was continued (250 rpm). After 6 h, the TCMS-coated glass slides were rinsed with n-hexane and dried under a nitrogen stream. To coat glass slides with a thin nanofilament film (1–2 µm thickness) for the chemical stability test, the reaction time was reduced to 3 h without stirring, and a water concentration of 180 ppm was used. To modify the TCMS-coated glass slides with PFDTS, they were activated in an oxygen plasma chamber (Diener Electronic Femto, 120 W, 2 min, 7 cm3 min-1 oxygen flow rate). Subsequently, 180 μL of PFDTS was mixed with 350 mL of n-hexane. The activated substrates were immersed in the solution for 20 min, rinsed with n-hexane, and dried under a nitrogen stream. ADVANCING, RECEDING AND ROLL-OFF ANGLE MEASUREMENTS Receding contact angles (𝜃𝜃Rec) and roll-off angles were measured for different liquids in order to characterize the gas contactor membranes. The receding contact angle denotes the lowest contact angle during the receding of a drop. On a rough surface, the receding contact angle is defined as the apparent, macroscopic contact angle of the liquid observed during receding on a length scale much larger than the scale of the microstructures. In many cases, the static contact angle depends on how a drop is deposited on a surface and can adopt every value between the receding and advancing contact angle.[29, 61, 278-283] Fur- thermore, even a wetted surface can show a high static contact angle.[29, 284] The receding contact angle enables an accurate comparison between different membranes, as it de- pends on the adhesion of the liquid to the substrate.[7, 87] If the liquid penetrates in the pores, the adhesion is high, resulting in a pinned contact line and low 𝜃𝜃Rec. On superhy- drophobic and superomniphobic surfaces, the advancing contact angle is close to 180°, as the advancing front needs to bend down to touch the next protrusion.[61, 283] Nonetheless, FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 101 advancing contact angles using a goniometer with a video camera were measured and summarized in Table S2. For transparent substrates, contact angles can be measured with much higher accuracy by laser scanning confocal microscopy (see Figure 38c). Advancing and receding contact angles were measured using a DataPhysics OCA35 con- tact angle goniometer (Table S2). Initially, 6 µL drops were deposited on the membranes. Afterward, 20 µL of the liquid was added to and removed from the drop. The measurement was consecutively repeated three times at the same spot and at three different spots per membrane. The error of the advancing and receding contact angle measurements was estimated to be ±4° (or greater) depending on the liquid and the membrane. For the superhydrophobic and commercial PTFE and PP membranes, the three-phase contact line of 20 wt% and 50 wt% aqueous MDEA solutions was strongly pinned. The contact line did not recede when removing 20 µL liquid. Especially the 50 wt% MDEA so- lution wetted these membranes. To determine the receding contact angle for the super- hydrophobic, PP and PTFE membranes 20 µL MDEA solution was added to the initial 6 µL droplet, and then 24 µL was removed. The contact line started to recede during the re- moval of the last few microliter (Movie S3). For the superhydrophobic and PP-200 mem- branes and 50 wt% MDEA solution, 20 µL MDEA (50 wt%) solution was added to the initial 6 µL droplet, and then 25.5 µL was removed. The contact line only started moving when almost the whole liquid was removed, again resulting in a low receding contact angle for MDEA (50 wt%) solution. Hexadecane instantly spread on superhydrophobic and all com- mercial membranes and diffused through the membranes (Movie S4). Therefore, the re- ceding contact angle could not be determined and was estimated to be below 5°. Roll-off angles were measured using the same DataPhysics OCA35 contact angle goniom- eter (Table S2). Therefore, 6 µL drops were deposited on the membranes, and the meas- uring plate was tilted until the drops rolled off. The roll-off angle was determined at least at 6 different spots per membrane, and mean roll-off angles and deviations were calcu- lated. SURFACE TENSION MEASUREMENTS Surface tensions of the aqueous MDEA solutions were determined with a DataPhysics DCAT11EC tensiometer using a cylindrical Wilhelmy plate (PT10, DataPhysics). Each sam- ple was measured three times. Average surface tension and standard deviations were cal- culated. 102 ENHANCING CARBON DIOXIDE CAPTURE USING ROBUST SUPEROMNIPHOBIC MEMBRANES CARBON DIOXIDE EXCHANGE EXPERIMENTS The home-built setup (Figure 39a) consisted of an ATR-FTIR spectrometer (Bruker Vertex 70 with external chamber XSA), a glass reservoir for the amine solution, a pump (Ismatec REGLO Digital) and the home-built exchange chamber (Figure 39b). The individual parts were connected using flexible Tygon tubes (Ismatec Tygon MHSL 2001) with an inner di- ameter of 1.5 mm. The spectrometer had a home-made flow cell with a height of 100 µm (Hellma Analytics). A germanium ATR-IR crystal was used (Korth-Kristalle). The measur- ing chamber was continuously purged with nitrogen to minimize the water and carbon dioxide content in the chamber. To measure the CO2 exchange, a membrane was put be- tween the upper and lower part of the exchange chamber and fixed in a holder. The flow channels of the exchange chamber had a width of 2 mm and a height of 1 mm (Figure 39b). The reservoir was filled with 6 mL aqueous MDEA (20 wt%) solution, and the pump was started (flow rate: 3.2 mL min-1, flow velocity: 2.7 cm s-1). The flow was split up before passing the spectrometer as the built-in flow cell had a height of 100 µm. Due to the large flow resistance, a 50 wt% MDEA solution could not be pumped through the flow cell, vis- cosity at 20 °C: 11.7 mPa∙s.[276] After the membrane was covered with the MDEA solution, the CO2 stream (1.5 mL min-1, 99.99%, Westfalen AG) valve was opened, and the FTIR measurement was started. FTIR-spectra were accumulated over a period of 30 s and aver- aged (≈120 spectra at a mirror frequency of 40 kHz) every 1.5 min until the measurement was stopped. The obtained spectra were analyzed using the Bruker OPUS 7.5 software. Therefore, the baseline of all spectra was corrected to take into account the drift of the baseline. The peak height/absorbance at 1302 cm-1 (bicarbonate band) was determined. The peak height was converted into the number of moles CO2 absorbed, as described in Experimental Section 4.4.12 Estimation of CO2 absorption and CO2 absorption flux. SCANNING ELECTRON MICROSCOPY Scanning electron microscopy (SEM) images were taken with a Zeiss LEO 1530 Gemini SEM at gun voltages of 1.5–3 kV using the in-lens detector. To avoid charging, samples were sputtered with 5 nm Pt before measurement using a BalTec MED 020 modular high vacuum coating system (with an argon pressure of 2 × 10−5 bar and a current of 60 mA). PORE CHARACTERIZATION The samples were dried at 70 °C in vacuo for 20 h. Nitrogen adsorption-desorption iso- therms at 77 K were measured using a Quantachrome Autosorb-1. BJH (Barrett-Joyner- FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 103 Halenda) analysis was used to determine the pore size distribution.[285] BET (Brunauer– Emmett–Teller) analysis was used to determine the BET surface area.[286] For the mercury intrusion porosimetry, the samples were dried at 70 °C for 19 h. Mercury intrusion porosimetry was measured using a Quantachrome Poremaster 60 GT. The Washburn equation was used to determine the pore size corresponding to the applied pressure. Gas pycnometry (in accordance with DIN 66137) using helium was measured using a Quantachrome Ultrapycnometer. LASER SCANNING CONFOCAL MICROSCOPY Laser scanning confocal microscopy (LSCM) images were taken using Leica LAS X soft- ware and an inverted confocal microscope (Leica TCS SP8 SMD, HCX PL APO 40x/0.85 CORR CS dry objective) with a resolution of about 0.3 and 1.0 μm in the horizontal and vertical direction, respectively. A superomniphobic glass slide was placed in the sample holder of the microscope, and an MDEA (50 wt%) droplet stained with ATTO488 dye was placed on the glass slide. The surface tension of the MDEA (50 wt%) drop did not change due to the addition of the dye (concentration: 1 mg L-1). The fluorescence of ATTO488 in the MDEA solution is shown in pink, and the reflection of light from the air-MDEA solu- tion interface is shown in blue. Reflection and emission signals were recorded simultane- ously. TENSILE STRENGTH MEASUREMENTS Tensile specimens (4 mm width) were punched out of the superomniphobic and the com- mercial membranes. The thicknesses of the superomniphobic, the PTFE (220 nm and 1 µm avg. pore size) and the PP membranes were 160 µm, 220 µm, 200 µm, and 170 µm, respectively. The PTFE membranes had a high-density polyethylene supporting mesh at the backside. The tensile specimens were fixed in the holder of Zwick/Roell Z005 univer- sal testing machine. A preload of 0.05 N was applied, and the samples were extended at 10 mm min-1 until rupture. Force-strain curves were recorded. ESTIMATION OF CO2 ABSORPTION AND CO2 ABSORPTION FLUX First, the CO2 loading of a CO2-saturated aqueous 20 wt% MDEA solution at 20 °C and stand- ard pressure (1 atm) was determined by bubbling CO2 in 3.385 g MDEA solution for 2.5 h. The mass increase caused by the uptake of CO2 was Δ𝑚𝑚CO2 = +0.157 g. Since a small amount of the solution evaporates, the same experiment was repeated using N2 instead of 104 ENHANCING CARBON DIOXIDE CAPTURE USING ROBUST SUPEROMNIPHOBIC MEMBRANES CO2; Δ𝑚𝑚N2 = -0.074 g. The decrease in mass caused by evaporation was added to the in- crease in mass due to CO2 uptake, Δ𝑚𝑚total = 0.231 g. Using the molecular weight of CO2 (44.01 g mol-1), we found that 5.25 mmol of CO2 was absorbed by the MDEA solution. As a result, the loading of a CO2-saturated MDEA (20 wt%) solution was 𝛼𝛼Load ≈ 0.93 mol CO2 per mol MDEA. This result is in good agreement with previously published data (𝛼𝛼Load = 0.786 for 20.5 wt% MDEA solution at 50 °C and 106 kPa CO2 partial pressure,[267] 𝛼𝛼Load = 0.866 for 23.4 wt% MDEA solution at 40 °C and 101 kPa CO2 partial pressure,[33] higher tem- peratures decrease the loading capacity).[267, 287] The average CO2 absorption flux during the first 30 min of gas exchange (using a super- omniphobic membrane) was estimated by measuring the peak height of the carbonate band at 1302 cm-1 after 30 min (average peak height based on three individual measure- ments). During the first 30 min, the peak height increased almost linearly. The saturation ratio was calculated by dividing the peak height after 30 min by the peak height at satura- tion: 𝑆𝑆r = 0.0051/0.0162 = 0.31 (= 31%). Knowing the volume of the MDEA solution and the concentration (6 mL, density 𝑔𝑔 = 1.0169 g cm-3 at 20 °C),[23] the number of moles of MDEA in the solution was calculated to be 10.2 mmol. Using the loading ratio (𝛼𝛼Load ≈ 0.93 mol CO2 per mol MDEA) of a saturated MDEA solution and 𝑆𝑆r = 0.31, we found that 2.9 mmol of CO2 were transferred through the membrane and absorbed by the solution during the first 30 min. Thus, over an exchange area of 𝐴𝐴 = 6.36 cm2 of our exchange chamber, the CO2 absorption flux was calculated to be 𝐽𝐽 = 2.5 mmol m-2 s-1 for the super- omniphobic membrane. AUTHOR CONTRIBUTIONS F. Geyer fabricated the superomniphobic membranes, carried out the experiments, and wrote the manuscript. C. Schönecker and F. Geyer developed the setup for CO2 quantifi- cation. F. Geyer, C. Schönecker, D. Vollmer, and H.-J. Butt contributed to experimental planning, data analysis, and manuscript preparation. All authors reviewed and approved the manuscript. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 105 ACKNOWLEDGMENTS We thank for stimulating discussions, for technical support and for carefully reading the manuscript. We thank for the graphical realization of Figure 39. We thank for nitrogen isotherm measurements. This work was supported by the ERC advanced grant 340391-SUPRO and COST MP1106. 106 HOW TO COAT THE INSIDE OF NARROW AND LONG TUBES WITH A SUPER-LIQUID-REPELLENT LAYER 5 HOW TO COAT THE INSIDE OF NARROW AND LONG TUBES WITH A SUPER-LIQUID-REPELLENT LAYER — A Promising Candidate for Antibacterial Catheters Florian Geyer,1 Maria D’Acunzi,1 Ching-Yu Yang,1 Michael Müller,1 Philipp Baumli,1 Anke Kaltbeitzel,1 Volker Mailänder,1 Noemí Encinas,1 Doris Vollmer,1 and Hans-Jürgen Butt1 1Max Planck Institute for Polymer Research, Ackermannweg 10, 55128 Mainz, Germany Abstract: Fouling of thin tubes is a major problem, leading to various infections and as- sociated morbidities while cleaning is difficult or even impossible. Here, a generic method is introduced to activate and coat the inside of meter-long and, at the same time, thin (down to 1 mm) tubes with a super-liquid-repellent layer of nanofilaments, exhibiting even antibacterial properties. Activation is facilitated by pumping an oxidative Fenton so- lution through the tubes. Subsequent pumping of a silane solution renders the surface of the tubes super-liquid-repellent. The wide applicability of the method is demonstrated by coating stiff and flexible tubes made of polymers, inorganic/organic hybrids, metals, and ceramics. Coated medical catheters show excellent antibacterial properties. Notably, the nanofilaments retain their antibacterial properties even in the superhydrophilic state. These findings open new avenues toward the design of biocide-free, antibacterial tubings and catheters. Keywords: wetting, superhydrophobic, superoleophobic, medical, anti-biofouling This chapter is reprinted with permission from F. Geyer et al., Adv. Mater. 2019, 31, 1801324. Copyright 2019, John Wiley and Sons The Supporting Information is available from Wiley Online Library: https://doi.org/10.1002/adma.201801324 FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 107 INTRODUCTION Pipes, tubes, capillaries, and hoses (in the following summarized as tubes) of different shapes and materials are essential for our daily life and numerous industrial processes. For example, thin tubes are used for food processing, water filtration[288], gas separa- tion,[173] microfluidics[289], or as an insulating material. In medicine, the use of polymeric tubes is inevitable, for example, for infusions or catheters.[290] An ideal coating permits to tune the flow,[31, 289] the wetting properties,[1, 3-4, 291] and to suppress biofilm for- mation.[165-166, 292] As cleaning of thin tubes is challenging or even impossible, the develop- ment of improved coatings is a prominent issue. Some super-liquid-repellent layers can effectively reduce or delay biofilm formation, which is the irreversible attachment and proliferation of bacteria,[159-160, 162] thereby diminishing the risk of infection-related prob- lems (failure and replacement of implants or devices, and even sepsis). Thus, a promising strategy might be to coat the inside of tubes with a super-liquid-repellent layer. Applying such a coating may suppress biofilm formation and thus save the lives of thousands of patients in hospitals and retirement homes through the reduction of pathogenic infec- tions. Although several methods exist to coat an open surface with a super-liquid-repel- lent layer,[30, 82, 84, 293-294] to the best of our knowledge, no technique is described that allows coating the inside of meter-long and narrow medically relevant tubes (diameter down to 1 mm). Here, we introduce a generic method to overcome this challenge. Characteristic for super-liquid-repellent coatings are their outstanding self-cleaning properties.[31, 145, 153] When placing drops of water or complex liquids like blood on such a surface, the drop takes an almost spherical shape and rolls off when the surface is tilted by less than 10°.[41, 82, 95] To render a surface super-liquid-repellent, it needs to be coated with nano- or micrometer-sized protrusions with overhanging morphology and low sur- face energy.[87, 295-297] Such topology generates an energy barrier for the liquid so that water drops rest on the protrusions entrapping air underneath. As a result of this so-called Cas- sie state,[27] the liquid is more in contact with air rather than with a solid, leading to high apparent contact angles.[28, 298] Current methods to generate superhydrophobic and super- liquid-repellent coatings on non-tubular substrates such as electrospinning,[82] lithogra- phy,[42] dip coating,[38] spin coating,[144] or spray coating[40, 299] are unsuitable for tubes hav- ing an inner diameter of only a few millimeters. Chemical etching methods of metal pipes, for example, copper,[102] can produce microroughness. Nonetheless, etching is limited to the respective metal, and metal tubes are rigid and prone to corrosion. Other techniques 108 HOW TO COAT THE INSIDE OF NARROW AND LONG TUBES WITH A SUPER-LIQUID-REPELLENT LAYER require activation of the surface, a challenge for long and narrow polymeric tubes. Pira- nha cleaning solutions (a mixture of sulfuric acid, water, and hydrogen peroxide) are suit- able for glass surfaces but cannot effectively activate polymeric surfaces. Usual plasma treatments are only applicable to activate short tubes as the plasma cannot be inducted inside of high aspect ratio tubes. Being able to coat the inside of high aspect ratio tubes will extend the range of applications of super-liquid-repellent surfaces.[102, 147] Here, we develop a universal route to activate and subsequently coat the inside of meter- long and, at the same time, thin tubes (down to 1 mm) with a super-liquid-repellent layer of nanofilaments. Activation of long tubes is enabled by making use of the Fenton reac- tion, which can be pumped through the tubes. The wide applicability of the coating method is demonstrated for tubes consisting of various materials, including polymers, hybrid materials, metals, and ceramics. Excellent anti-biofouling properties are demon- strated using nanofilament-coated polyurethane catheters used in hospitals. RESULTS AND DISCUSSION To coat the inside of the tubes with a super-liquid-repellent layer of nanofilaments, the surface needs to be hydrophilic, having polar functional groups, preferentially hydroxy groups. Thus hydrophobic, unfunctionalized polymer surfaces need to be activated. Short polymer tubes can be activated via oxygen plasma because the ionizing plasma can suffi- ciently penetrate from the ends of the tube. Plasma activation leads to the formation of various functional groups like hydroxy, carboxyl, carbonyl, and peroxide.[300] After activa- tion, the standard procedure (Figure 42a) to coat tubes with a layer of silicone nanofila- ments is to immerse the hydrophilic tubes in a solution of trichloromethylsilane (TCMS), in toluene or n-hexane for example, in the presence of trace amounts of water (85– 225 ppm). After a reaction time of at least 1 h, the surface is covered with a layer of nano- filaments having a thickness of a few micrometers. During the reaction, the TCMS hydro- lyzes due to water traces in the solvent. The hydrolyzed TCMS molecules react with hy- droxy groups on the surface and induce a polysiloxane (i.e., silicone) polymerization on the surface of the tube (Figure S1 summarizes the involved chemical reactions). The coat- ing is superhydrophobic as the methyl groups of the polysiloxane orient to the outside to lower the surface energy.[67, 112, 115-116] FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 109 - Figure 42. Illustration of the coating procedure, cross-sectional optical images, and scanning electron microscopic (SEM) images of the inner surface of the nanofilament-coated tubes. a) After activation, the nanofilament coating can be applied by immersing in or flushing the tubes with a trichloromethylsilane solution for a certain reaction time. SEM images demonstrate the formation of nanofilaments. Subse- quent activation and fluorination yield a super-liquid-repellent coating. The different colors mark the change of the surface chemistry or topography of the interior surface. The photographs of the drops illus- trate the changes in the wetting properties after each individual step on a flat surface. b) High-density polyethylene (HDPE) tube. c) Polypropylene (PP) tube. d) Low-density polyethylene (LDPE) tube. e) Poly- urethane (PU) tube. f) Polyurethane/barium sulfate (PU/BaSO4) tube. g) Brass tube. h) Alumina (Al2O3) ce- ramic tube. i) Zirconia (ZrO2) ceramic tube. All SEM images show the formation of nanofilaments having diameters between 10–100 nm on the inner surface of the tubes. Scale bars: a) 500 nm; b–i) 500 nm for SEM images; 1 mm for optical images. To render the coating super-liquid-repellent toward surfactant solutions or blood, the nanofilament-coated tubes are activated and modified with a perfluoro silane (1H,1H,2H,2H-perfluorodecyltrichlorosilane; see Figure S2 for XPS spectra of as-synthe- sized, plasma-activated and fluorinated nanofilaments in a tube). 110 HOW TO COAT THE INSIDE OF NARROW AND LONG TUBES WITH A SUPER-LIQUID-REPELLENT LAYER Notably, fluorinated nanofilaments can also be achieved in a one-step reaction[123] as ver- ified by coating high-density polyethylene with filaments (Figure S3). Here, a mixture of tetraethoxysilane (TEOS) and 1H,1H,2H,2H-perfluorodecyltrichlorosilane (PFDTS) in tol- uene was used, containing trace amounts of water (100 ± 5 ppm). TEOS and PFDTS hydro- lyze due to the water traces in solution, inducing a polymerization on the surface of the tube, co-assembling into nanofilamentous structures. The wide applicability of the method is demonstrated by coating tubes of varied materials having diameters between 1 mm and 3 mm. We applied the coating on commonly used and medically relevant polymer tubes, ranging from high-density polyethylene (HDPE), polypropylene (PP), low-density polyethylene (LDPE), and polyurethane (PU) to or- ganic/inorganic hybrid polyurethane/BaSO4 (Figure 42b–f). The PU and PU/BaSO4 tubes are used in hospitals as catheter tubes, with BaSO4 as an X-ray contrast medium. In addi- tion, the coating was applied to rigid brass, alumina (Al2O3), zirconia (ZrO2), and copper (Cu) tubes (Figure 42g–i; Figure S10 for Cu. The HDPE, PP, LDPE, and ceramic tubes were coated using TCMS (34 mmol L-1) dissolved in toluene with 135 ± 10 ppm water content for 3 and 20 h, respectively. The brass and copper tubes were coated using a water content of 225 ± 5 ppm for 3 h. The polyurethane tubes (PU and PU/BaSO4) were coated in a water-saturated (85 ± 5 ppm water content) TCMS (17 mmol L-1) solution in n-hexane for 20 h, showing that nanofilaments grow in both, toluene and n-hexane. Scanning electron microscopic (SEM) images demonstrate the formation of nanofilaments on all tubes. On the HDPE, PP, and LDPE tubes, a coating consisting of mainly thinner nanofilaments (10–30 nm diameter) and some lone thicker nanofilaments (40–70 nm diameter) was obtained (Figure 42b–d). On the polyurethane tubes, the diameter of the nanofilaments varied between 20 and 70 nm in diameter. In addition, silicone nanoparticles between 150 and 500 nm diameter were found inside the coating (Figure 42e,f). On brass, the coating consisted of nanofilaments with ≈50 nm di- ameter (Figure 42g). Ceramic tubes showed rather short and thin nanofilaments between 15 and 40 nm in diameter (Figure 42h,i). More detailed SEM images of the bare and coated tubes are shown in Figure S4–S12. The necessity of activation is demonstrated for LDPE tubes as a model system (Figure 43). Without activation, only a few randomly distributed nanofilaments formed on LDPE (Figure 43a). After plasma activation, a short tube was coated with a homogeneous layer of nanofilaments (Figure 43b). However, plasma activation is only applicable inside of FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 111 short tubes because of limitations of the flux to penetrate longer tubes. Thus, a more ver- satile method to activate the inner surface of tubes of arbitrary length is required. Here, we introduce the Fenton reaction[251] for activating the surfaces. The Fenton reaction is industrially widely used to oxidize contaminants or to treat wastewater.[252] Fenton’s rea- gent is an aqueous solution of hydrogen peroxide and ferrous iron. The ferrous iron acts as a catalyst to produce hydroxyl radicals and hydroxide ions as a byproduct: Fe2+ + H2O2 → Fe3+ + HO· + OH- These hydroxyl radicals have the second-highest oxidation potential (redox potential 2.80 V) of all common oxidation agents, only surpassed by fluorine (redox potential 3.03 V). Similar to the oxygen plasma, the hydroxyl radicals oxidize the polymeric surface inside the tubes introducing functional groups containing oxygen, like hydroxy, peroxide, and carboxyl (Figure 43c). X-Ray photoelectron spectroscopy (XPS) spectra of a bare and Fenton-activated LDPE tube revealed an increase of oxygen-containing functional groups (Figure 43d). This is manifested in the increase of the ratio between the oxygen (O1s) peak to the carbon (C1s) peak. The structure of the nanofilament coating depends on the acti- vation time with Fenton’s reagent and thus on the density of oxygen-containing functional groups on the polymer surface. Already after 30 min of exposure to a Fenton solution (22 mmol L-1 FeSO4, 0.50 mol L-1 H2O2, pH = 2.8, cooled in an ice bath) under a flow rate of 8 mL min-1, a coating featuring mostly thick nanofilaments resembling particles chains (>100 nm) could be obtained on an LDPE tube (Figure 43e and Figure S13). After 60 min exposure, the coating was composed of filaments with an average diameter below 100 nm. After 90 min, the average nanofilament diameter decreased to ≈50 nm. After 120 and 180 min exposure to Fenton’s solution, the nanofilaments were thicker than 50 nm. In addition, more nanoparticles started to form, which were mostly assembled into particle chains. To coat a 2 m long and 1 mm thick LDPE tube, the inner surface was activated by pumping Fenton’s reagent (22 mmol L-1 FeSO4, 0.50 mol L-1 H2O2, pH = 2.8, 90 min, cooled in an ice bath) through the tube at a flow rate of 8 mL min-1 (see Experimental Section 5.4.7 for further details). After Fenton-activation, a TCMS-solution (34 mmol L-1 in toluene, 150 ± 5 ppm water content) was slowly pumped (0.1 mL min-1) through the tube to ensure a homogeneous coating and to prevent the depletion of silane within the tube. 112 HOW TO COAT THE INSIDE OF NARROW AND LONG TUBES WITH A SUPER-LIQUID-REPELLENT LAYER Figure 43. Activation and coating of a two-meter long tube. a) SEM image of low-density polyethylene (LDPE) tube coated without prior activation. b) SEM image of a short tube activated using oxygen plasma before the coating was applied. c) Scheme of the activation of polymer surfaces using the Fenton reagent. d) X-ray photoelectron spectroscopy (XPS) spectra of a bare and Fenton-activated LDPE tube. e) SEM im- ages of coated Fenton-activated LDPE tubes after different Fenton reagent exposure times using a flow rate of 8 mL min-1. f) SEM images of a coated LDPE tube with a total length of 2 m and an inner diameter of 1 mm at both endings and the middle of the tube. The formation of nanofilaments could be observed at all positions. Scale bars: a, b) 500 nm; f) 1 µm. Indeed, SEM images, taken at both endings of the tube and in the middle of the tube, doc- ument the formation of nanofilaments along the inside of the tube having an aspect ratio of 2000:1 (Figure 43f and Figure S14). Due to the ability to pump the Fenton solution and the coating solution through the tube, the whole coating procedure can be applied to FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 113 tubes of greatly varying materials and aspect ratios. The Fenton reaction can also be uti- lized to activate the nanofilaments and thus replace the plasma activation before fluori- nation of the nanofilaments (Figure S15a,b). A nanofilament-coated tube was hydrophi- lized by pumping the Fenton solution through the tube for 50 min at a flow rate of 8 mL min-1. Afterward, water drops easily spread inside the tube (Figure S15c,d). The wetting properties of a bare PU/BaSO4 tube, fluorinated nanofilaments on polyure- thane tubes, and fluorinated nanofilaments on the 2 m-long LDPE tube were investigated using different liquids, including liquids relevant for biomedical applications (Figure 44; Figure S16 and S17). Due to the small inner diameters (1.3 and 0.95 mm) and the high curvature of the tubes, contact angles cannot be determined in a standard way, i.e., using the goniometer technique. However, super-liquid-repellency is typically accompanied by only a few pinning sites and low adhesion of drops to the surface. We quantified the wet- tability indirectly by monitoring the adhesion of drops to the coating. Therefore, the PU tubes were horizontally cut in the middle prior to coating in order to avoid pinning sites at the cut surface. Drops between 1 and 3 µL volume were placed onto the inside of the tubes. Next, the syringe was slowly moved up and down (see also Movie S1–S4). On bare PU/BaSO4 tubes, the high adhesion between the phosphate-buffered saline (PBS, surface tension, 𝛾𝛾 = 72.8 mN m-1) and PU/BaSO4 caused that the drop detached and stuck at the surface (Figure 44a). Contrary, on the PU/BaSO4 tubes coated with fluorinated nanofila- ments, drops of PBS, urine, and blood plasma hardly showed any adhesion (Figure 44b– d). The drops could be easily removed by retracting the syringe. Drops of diiodomethane (CH2I2, 𝛾𝛾 = 50.8 mN m-1) detached from the syringe upon retraction due to the high den- sity of CH2I2 (3.3 g mL-1) and rolled out of the tube (Figure 44e). Also, common food liquids like water, Cola, apple juice spritzer, and the non-polar aromatic liquid iodobenzene (𝛾𝛾 = 39.7 mN m-1) showed low adhesion and could be easily removed from a coated PU tube (Figure S16). The low adhesion of CH2I2 and iodobenzene drops showed the super liquid- repellency of the coating also to non-polar liquids. The super liquid-repellency of the coated 2 m-long LDPE tube was confirmed at both endings and in the middle (Figure S17). Video microscopy showed that all aqueous drops exhibited a lower adhesion than 3 µL water drops on a superhydrophobic micropillar array (rectangular, 10 µm height with 5 × 5 µm2 top areas; pillar–pillar distance (center to center): 20 µm; 9° ± 1° roll-off angle for 6 µL water drops). On the micropillar array, the drop either stuck to the surface or its detachment from the surface caused visible vibrations of the drop (Movie S5). As a refer- ence, we also measured the receding contact angles and roll-off angles on nanofilament- 114 HOW TO COAT THE INSIDE OF NARROW AND LONG TUBES WITH A SUPER-LIQUID-REPELLENT LAYER coated glass slides and on flat polyurethane/BaSO4. The tube material was flattened by heating. Phosphate buffered saline on flat PU/BaSO4 showed a receding contact angle of 43° ± 4° and did not roll off. However, PBS on a nanofilament-coated glass slide showed a receding contact angle (RCA) of 158° ± 4° and a roll-off angle of 2° ± 1° for 6 µL drops. After plasma or Fenton activation of the nanofilaments, the receding contact angle of PBS was ≈0°, due to spreading. On a glass slide coated with fluorinated nanofilaments, PBS and CH2I2 as an example showed a receding contact angle of 160° ± 4° and 149° ± 4°, respectively, and roll-off angles of 1° ± 1° and 1° ± 1° for 6 µL drops, respectively. The extremely low adhesion of the drops to the coated inner surfaces of the tubes is in agreement with the high receding contact angles and low roll-off angles measured on nanofilament-coated glass slides. Figure 44. Wetting properties of a coated and fluorinated PU/BaSO4 catheter tube in com- parison to a bare PU/BasO4 tube. a) A 2 µL PBS (phosphate-buffered saline) drop sticking on a bare tube. b,c) A 2 µL PBS and a 3 µL urine droplet on a coated tube; both showed very low adhesion and could be easily removed by retracting the syringe. d) Even a 2 µL blood plasma drop showed very low adhesion to a coated tube and was easily removed from it. e) A 2 µL diiodomethane drop on a coated tube; during retraction of the syringe, the droplet detached from the syringe due to the high density of diiodomethane and rolled out of the tube. Scale bars: 0.5 mm. Many super-liquid-repellent coatings suffer from poor mechanical stability or easy de- lamination.[301] To test the long-term stability of the coating against flow-induced shear, water was pumped through a nanofilament-coated LDPE tube (0.95 mm inner diameter) for up to 10 months. The flow was adjusted to the maximum accessible pumping speed of the peristaltic pump (11 mL min-1), which corresponds to an average flow velocity of FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 115 ≈30 cm s-1 and a shear rate at the tubes’ inner surface of |𝛾𝛾̇ | ≈ 2 × 103 s-1 and a shear force per unit area of ?̇?𝛾η = 2 N m-2 (here, η = 0.001 Pa·s is the viscosity of water; see Experimental Section 5.4.15 and Figure 46 for further details). The morphology of the coating was checked after 5, 10, 30 d, and 2, 3, and 10 months (Figure S18). Surprisingly, even after 10 months, SEM images show that the morphology remained unaltered, which demonstrates the excellent adhesion of the nanofilaments to the substrate and the high stability of the coating. Additionally, we exposed fluorinated nanofilaments coated on a glass slide to an even higher flow velocity of >220 cm s-1 by impacting a water jet at an angle of ≈35° for 1 h (Figure S19a,b). SEM analysis revealed that the coating was not abraded, delaminated, or destroyed but remained unaltered (Figure S19c,d). The roll-off angle of 6 µL water drops stayed at 1° ± 1°, confirming the high stability of nanofilaments against liquid streams. Furthermore, tubes are typically subjected to bending, and thus the coating needs to be flexible. Therefore, we tested the flexibility and resistance of the coating toward bending. The coating did not alter, delaminate, or break even after 1000× bending cycles (Figure S20). The methyl groups confer flexibility to the siloxane polymer. It has been shown that superhydrophobic and liquid-repellent coatings may effectively delay biofilm formation.[159-160, 162] Superhydrophobicity leads to very high contact angles of the bacterial suspension on the surfaces as well as a low solid-liquid contact area. This translates into few adhesion sites (Cassie state) that can be accessed by the bacteria, re- sulting in low bacterial adhesion. Inspired by these results, we evaluated the potential of coated PU tubes regarding their anti-biofouling properties for medical devices.4 There- fore, we coated and incubated PU tubes (1 cm) in concentrated E. coli bacteria solutions (108 cells mL-1) for 168 h at 37 °C. The samples included bare PU catheters as a reference, as-prepared nanofilament-coated PU tubes (Me-NF), superhydrophilic nanofilament- coated PU tubes (OH-NF) and fluorinated nanofilament-coated PU tubes (F-NF) (Figure 45). After incubation, the inside of the bare PU tubes was homogeneously covered with E. coli throughout the whole tube (Figure 45a). On the nanofilament-coated tubes, the coverage with E. coli was greatly reduced. Hardly any E. coli could be found (Figure 45b–d). The squares mark positions where a single bacterium was located. The E. coli bacteria had a few micrometers in length and adhered to the top faces of a few nanofilaments (Figure 45b, right). On the as-prepared and fluorinated nanofilaments, the 4During the publication of this work, also another research group investigated the antibacterial properties of superhydrophobic nanofilaments (coated on glass slides).[164] 116 HOW TO COAT THE INSIDE OF NARROW AND LONG TUBES WITH A SUPER-LIQUID-REPELLENT LAYER number of adhered bacteria was reduced by more than 20 times. Notably, on the super- hydrophilic nanofilaments, the number of bacteria per cm2 was reduced by up to almost two orders of magnitude, 85 times (Figure 45e, see Note S1 in the Supporting Information for an in-depth discussion on the role of superhydrophobicity, Cassie state, and cell mo- tility on bacterial adhesion). Figure 45. Antibacterial properties of a bare PU tube, a nanofilament-coated PU tube, a superhydrophilic nanofilament- coated PU tube, and a fluorinated nanofila- ment-coated PU tube after an incubation time of 168 h in an E. coli solution. a) Bare PU tube showing high E. coli coverage. Scale bars: 20 µm (left); 4 µm (right). b) Nanofilament-coated PU tube showing highly reduced bacterial coverage (left). High magnification of a single E. coli bacte- rium attached to nanofilaments (right). Scale bars: 4 µm (left); 500 nm (right). c) Superhydrophilic nanofilament-coated PU tube. Scale bar: 4 µm. d) Fluorinated- nanofilament-coated PU tube. Scale bar: 4 µm. e) Average number of bacteria per area adhered to the surface for the different tubes after 168 h of incubation. Thus, the presented coating is capable of effectively reducing bacterial adhesion and bio- film formation in the superhydrophobic as well as in the fully wetted state, the so-termed FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 117 Wenzel state. We attribute this excellent anti-biofouling property to the characteristic spacing of the irregular three-dimensional layer of nanofilaments, which falls just below the size of E. coli (Figure 45b). Therefore, the stiff bacteria cannot enter the pores. Fur- thermore, the small diameter of the nanofilaments and their irregular three-dimensional arrangement reduces the anchoring area. This distinctive combination of length scales and morphology makes tubes coated with nanofilaments a highly promising candidate for medical applications, for example, as a catheter. CONCLUSION In conclusion, pumping Fenton’s solution through thin polymeric tubes offers a versatile and simple method to activate their inner surfaces. The activated surfaces can subse- quently be coated with a layer of nanofilaments. Due to the ability to pump both the Fen- ton solution and the coating solution, the wetting properties of meter-long and at the same time thin tubes (here: 2 m with 1 mm diameter) or other bicontinuous surfaces can be tuned from superhydrophilic to super-liquid-repellent. The Fenton activation may also be used to apply other liquid-repelling or antibacterial coatings, for example, liquid-like PDMS[68] coatings. Pumping offers the possibility that the activation and the coating step can easily be parallelized. Hundreds of parallel arranged tubes can be treated simultane- ously. Coated medical catheters prove to be promising candidates for reducing bacterial growth. Remarkably, even in the superhydrophilic state, the attachment of bacteria is highly reduced. Thus, the irregular 3D topography of nanofilaments greatly reduces the anchoring area for bacteria. Besides the excellent antibacterial properties, the coating can also withstand high flow velocities, bending, and does not delaminate. The wide ap- plicability of the nanofilament coating on tubes of various diameters and materials makes this method useful for a broad range of applications. It opens new avenues to fabricate biocide-free coatings of medical devices since the reduced bacterial adhesion and prolif- eration reduce the infection risk of patients. EXPERIMENTAL SECTION MATERIALS If not stated otherwise, experiments were carried out at room temperature. The following chemicals were used: trichloromethylsilane (TCMS, 99%, Sigma-Aldrich), tetraethoxy- silane (TEOS, 98%, Sigma-Aldrich), 1H,1H,2H,2H-perfluorodecyltrichlorosilane (PFDTS, 118 HOW TO COAT THE INSIDE OF NARROW AND LONG TUBES WITH A SUPER-LIQUID-REPELLENT LAYER 96%, Alfa Aesar), SU-8 3010 photoresist (MicroChem), mr-Dev 600 developer (micro resist technology), n-hexane (99.99%, Fisher Chemical), toluene (99.5%, VWR), ethanol (99.8%, Sigma-Aldrich), hexadecane (99%, Sigma-Aldrich), diiodomethane (99%, Sigma-Aldrich), iodobenzene (Sigma-Aldrich, 98%), sulfuric acid (99.99%, Sigma-Aldrich), iron sulfate heptahydrate (>99%, Sigma-Aldrich), hydrogen peroxide (35 wt%, VWR), phosphate- buffered saline (PBS, Gibco), glutaraldehyde (25 wt%, Sigma-Aldrich). The culture medium, LB (Lysogeny broth) medium, supplemented with Ampicillin (Cat. fas-am-b), was purchased from InvivoGen. Reagents were used as received. Cola and apple juice spritzer (60% apple juice concentrate) was obtained from the local supermarket. Whole blood was taken at the Department of Transfusion Medicine of the University Medical Center of the JGU Mainz from healthy donors. The study was approved by the local ethics committee and is in accordance with the Declaration of Helsinki. Sodium citrate was added to prevent clotting, and plasma was separated by centrifugation. The plasma was pooled, and aliquots were stored at -80 °C. A protein concentration of 65–70 g L-1 was determined by protein assay Pierce 660 nm. Urine was provided by F. Geyer. Purified and deionized water was obtained from a Sartorius Arium 661 VF water purification system and used for all experiments. Polypropylene (PP, inner diameter (ID): 1.6 mm, outer diameter (OD): 3.2 mm), high-density polyethylene (HDPE, ID: 2.0 mm, OD: 4.0 mm), and low-density polyethylene (LDPE, ID: 0.95 mm, OD: 2.0 mm) were obtained from RCT Reichelt Chemietechnik. Polyurethane (PU, ID: 1.3 mm, OD: 2.0 mm) and hybrid polyurethane/BaSO4 (PU/BaSO4, ID: 1.4 mm, OD: 2.1 mm) tubes were obtained from B. Braun. Ceramic tubes (Al2O3, ID: 1.5 mm, OD: 2.0 mm; ZrO2, ID: 3.0 mm, OD: 6.0 mm) were obtained from Friatec AG. Diameters are based on manufacturer specifications (except ID of LDPE). Brass (ID: 2.9 mm, OD: 5.0 mm) and copper (Cu, ID: 2.9 mm, OD: 4.0 mm) tubes were obtained from a local hardware store. Glass slides of 24 × 60 mm2 were obtained from Menzel-Gläser. Purified Milli-Q water from Sartorius Arium Pro VF was used for all experiments. NANOFILAMENT-COATED GLASS SLIDES Silicone nanofilaments on glass slides were synthesized by modifying previously reported methods.[112, 115] The fabrication contained the following steps:[116, 118, 233] Typically, glass slides were plasma-cleaned and activated in an oxygen plasma chamber (Diener Electronic Femto, 120 W, 5 min, 6 cm3 min-1 oxygen flow rate). TCMS was mixed with toluene having a water content of 180 ± 10 ppm (400 µL TCMS per 100 mL toluene). The FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 119 water content was evaluated using a Karl Fischer coulometer (Mettler Toledo C20 Compact KF coulometer). The solution was stirred for 60 s. Afterward, the activated glass slides were immersed in the solution, and the reaction chamber was sealed. After 6 h, the TCMS-coated glass slides were rinsed with n-hexane and dried under a nitrogen stream. Further details about nanofilament synthesis can be found elsewhere.[112, 115, 127] NANOFILAMENT-COATING OF THE HDPE, PP, LDPE, AND CERAMIC (AL2O3 AND ZRO2) TUBES USING PLASMA ACTIVATION Short HDPE, PP, LDPE, Al2O3 and ZrO2 tubes were plasma-cleaned and activated in an ox- ygen plasma chamber using different exposure times ranging from 36 s to 2 min and dif- ferent powers ranging from 30 W to 120 W (Diener Electronic Femto, 6 cm3 min-1 oxygen flow rate). TCMS was mixed with toluene having a water content of 135 ± 10 ppm (400 µL TCMS per 100 mL toluene). The water content was evaluated using a Karl Fischer coulom- eter (Mettler Toledo C20 Compact KF coulometer). The solution was stirred for 60 s. Af- terward, the tubes were immersed in the solution, and the reaction chamber was sealed. After 3 h (20 h for ceramic tubes), the TCMS-coated tubes were rinsed with n-hexane and dried under a nitrogen stream. NANOFILAMENT-COATING OF THE BRASS AND COPPER (CU) TUBES The brass and copper (Cu) tubes were cleaned by immersion in a 5 wt% sulfuric acid so- lution for 30 min. Afterward, the tubes were thoroughly rinsed with deionized water and activated in an oxygen plasma (Diener Electronic Femto, 120 W, 2 min, 6 cm3 min-1 oxygen flow rate). TCMS was mixed with toluene having a water content of 225 ± 5 ppm (400 µL TCMS per 100 mL toluene). The water content was evaluated using a Karl Fischer coulom- eter (Mettler Toledo C20 Compact KF coulometer). The solution was stirred for 60 s. Af- terward, the tubes were immersed in the solution, and the reaction chamber was sealed. After 3 h, the TCMS-coated tubes were rinsed with n-hexane and dried under a nitrogen stream. NANOFILAMENT-COATING OF THE PU AND PU/BASO4 TUBES The PU and PU/BaSO4 tubes were cleaned by flushing with an ethanol/deionized water mixture (50:50 to 40:60) for 5–10 min. The flushing also leads to the adsorption of a thin water layer, which facilitates nanofilament growth. TCMS was mixed with n-hexane hav- ing a water content of 85 ± 5 ppm (200 µL TCMS per 100 mL n-hexane). The water content 120 HOW TO COAT THE INSIDE OF NARROW AND LONG TUBES WITH A SUPER-LIQUID-REPELLENT LAYER was evaluated using a Karl Fischer coulometer (Mettler Toledo C20 Compact KF coulom- eter). The solution was stirred for 60 s. After flushing the tubes, they were immediately immersed in the solution, and the reaction chamber was sealed. Alternatively, the solu- tion was injected or pumped through the tube. After 20 h, the TCMS-coated tubes were rinsed with n-hexane and dried under a nitrogen stream. ONE-STEP FLUORINATED NANOFILAMENT-COATING ON HDPE TUBE The one-step synthesis of fluorinated nanofilaments is based on a method by Zhang et al.[123] The HDPE tube was plasma-cleaned and activated in an oxygen plasma chamber (Diener Electronic Femto, 6 cm3 min-1 oxygen flow rate, 150 W, 42 s). TEOS (100 µL) and PFDTS (16 µL) were mixed with toluene (80 mL) having a water content of 100 ± 5 ppm. The water content was evaluated using a Karl Fischer coulometer (Mettler Toledo C20 Compact KF coulometer). The solution was stirred for 60 s. Afterward, the tubes were im- mersed in the solution, and the reaction chamber was sealed. After 20 h, the coated tubes were rinsed with n-hexane and dried under a nitrogen stream. NANOFILAMENT-COATING OF TUBES USING FENTON REACTION ACTIVATION Iron(II) sulfate heptahydrate (0.25 g, 0.90 mmol) was dissolved in 40 mL of deionized wa- ter in a reaction vessel having an in- and outlet. The pH of the solution was adjusted to pH = 2.8 using a pH meter by adding sulfuric acid. The mixture was cooled for at least 30 min in an ice bath. The polymeric tube (LDPE) to be activated was connected to the outlet of the reaction vessel using a Tygon connection tube. The end of the polymeric tube was connected to a peristaltic pump (Ismatic Reglo Digital MS-2/12) and finally to the inlet of the reaction vessel using Tygon connection tubes. Hydrogen peroxide solution (1.8 mL, 35 wt%, 21 mmol) was slowly added to the acidic aqueous ferrous sulfate solution under vigorous stirring, and the mixture was stirred for 10 min. Afterward, the pump was started using a flow rate of 8 mL min-1. The Fenton[251] solution was pumped through the tube for the desired activation time. Typically, the activation time ranged between 20 min and 3 h. Every 15 min, the tube was emptied by reversing the pumping direction to eliminate gas bubbles, which may form on the inner surface due to the slight oxygen evolution during the Fenton reaction. After the desired activation time was reached, the tube was rinsed by flushing deionized water through the tube. Subsequently, the tube was dried by flush- ing dry nitrogen and further drying in vacuo. After drying, TCMS was mixed with toluene having a water content of 150 ± 5 ppm (400 µL TCMS per 100 mL toluene) in a reaction FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 121 vessel having an in- and outlet. The water content was evaluated using a Karl Fischer cou- lometer (Mettler Toledo C20 Compact KF coulometer). Similar to above, the Fenton- activated tube was connected to the reaction vessel and the pump. Here, Fluran F-5500-A tubes were used to connect the individual parts. After stirring the reaction mixture for 60 s, the reaction vessel was closed, the mixture was pumped into the activated tube, and the pump was stopped for 30 min. Subsequently, the pump was started again for 5 min with a flow rate of 0.1 mL min-1 and stopped again for 45 min. After that, the pumping process was continued for 20 h using a flow rate of 0.1 mL min-1. Subsequently, the TCMS- coated tubes were rinsed with n-hexane and dried under a nitrogen stream. PLASMA ACTIVATION OF NANOFILAMENT-COATED SURFACES AND TUBES To activate nanofilament-coated surfaces and short tubes, they were exposed to an oxy- gen plasma (Diener Electronic Femto, 120 W, 2 min, 7 cm3 min-1 oxygen flow rate). FENTON ACTIVATION OF NANOFILAMENT-COATED SURFACES AND TUBES To activate nanofilament-coated surfaces and tubes using Fenton reaction, iron(II) sulfate heptahydrate (2.5 g, 9.0 mmol) was dissolved in 10 mL of deionized water in a reaction vessel. The pH of the solution was adjusted to ≈3 by adding 10 µL of 10 wt% sulfuric acid. The nanofilament-coated substrate was immersed in ethanol, or ethanol was pumped through a nanofilament-coated tube to remove the air cushion (Cassie state) of the nano- filaments. Afterward, the substrate or tube was rinsed with deionized water to remove residual ethanol. The wet substrate was immersed in the acidic ferrous sulfate solution, and hydrogen peroxide solution (18 mL, 35 wt%, 0.21 mol) was slowly added (caution: strong reaction and gas formation occurs!) to the solution under vigorous stirring. After 10 and 20 min, additional H2O2 (each 18 mL, 35 wt%) was added. In the case of a tube, the solution mentioned above was pumped through the tube using a flow rate of 8 mL min-1. After 50 min, the substrate or tube was rinsed with 10 wt% sulfuric acid and copious amounts of deionized water. FLUORINATION OF ACTIVATED NANOFILAMENT-COATED TUBES OR SURFACES PFDTS was mixed with n-hexane (50 µL PFDTS per 100 mL n-hexane). The activated nan- ofilament-coated tubes or surfaces were immersed in the solution, and the reaction cham- ber was sealed. Alternatively, the solution was pumped through the tube. After 20 min the fluorinated tubes or surfaces were rinsed with n-hexane and dried under a nitrogen stream. 122 HOW TO COAT THE INSIDE OF NARROW AND LONG TUBES WITH A SUPER-LIQUID-REPELLENT LAYER SUPERHYDROPHOBIC SU-8 MICROPILLARS SU-8 micropillar arrays were prepared on glass slides by photolithography as previously reported.[62, 77] The rectangular pillars were designed 10 µm high with 5 × 5 µm2 top areas. The pillar–pillar distance between the centers of two adjacent pillars in a row was 20 µm. Details are described in Experimental Section 2.3.4. The SU-8 micropillars were activated in an oxygen plasma chamber (Diener Electronic Femto, 30 W, 2 min, 6 cm3 min-1 oxygen flow rate). Subsequently, PFDTS was mixed with n-hexane (50 µL PFDTS per 100 mL n-hexane). The activated SU-8 micropillar-coated glass slides were immersed in the solu- tion, and the reaction chamber was sealed. After 20 min the fluorinated pillars were rinsed with n-hexane and dried under a nitrogen stream. CONTACT ANGLE MEASUREMENTS AND DROP OBSERVATION Receding contact angles were measured using a DataPhysics OCA35 contact angle goni- ometer. Initially, 6 µL drops were deposited onto the substrate. Afterward, 20 µL of the liquid was added to and removed from the drop. The measurement was repeated at three different spots per substrate. The error of the receding contact angle measurements was estimated to be ±4°. Roll-off angles were measured using the DataPhysics OCA35 contact angle goniometer. Therefore, 6 µL drops were deposited on the substrates, and the measuring plate was tilted until the drops rolled off. The roll-off angle was determined in at least 6 different spots per substrate, and mean roll-off angles and corresponding standard deviations were calculated. The drop adhesion on the coated tubes was observed using the same DataPhysics OCA35 contact angle goniometer. Here, a syringe with a fluorinated tip was used. A 0.5–3 µL drop was generated at the tip and brought into contact with the coated tube. Afterward, the syringe was moved up and down to observe the adhesion of the drop to the tube. SCANNING ELECTRON MICROSCOPY Scanning electron microscopy (SEM) images were taken with a Zeiss LEO 1530 Gemini SEM at gun voltages of 3 kV using the in-lens detector. To avoid charging, samples were sputtered with Pt before measurement using a BalTec MED 020 modular high vacuum coating system (with an argon pressure of 2 × 10−5 bar and a current of 30 mA, 7 nm Pt). FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 123 X-RAY PHOTOELECTRON SPECTROSCOPY X-ray photoelectron spectra (XPS) were acquired using a Kratos Axis Ultra system. The X-ray source was monochromatic Al-Kα (source energy: 1486.69 eV). The take-off angle was 0°. The vacuum in the main chamber was below 4⋅10-10 mbar. The hybrid detector mode was used. Survey spectra were recorded with a pass energy of 160 eV. Charge neu- tralization was applied (current: 2.1 A, balance: 2.8 V, bias: 1.3 V). The survey scans were obtained with 2 scans. The software CasaXPS (Casa Software Ltd) was used for data elab- oration. The binding energies of the XPS spectra were calibrated with the help of the ali- phatic C1s peak at 285.0 eV. To measure XPS spectra inside of the tubes, the investigated tubes were cut in half horizontally and glued on carbon tape. CALCULATION OF THE SHEAR RATE: The flow of a Newtonian liquid through a long cylindrical tube (Figure 46) of radius 𝑟𝑟c is given by the Hagen-Poiseuille equation. d𝑉𝑉 𝜋𝜋∆𝑃𝑃𝑟𝑟4 = c (35) d𝑡𝑡 8 𝜂𝜂𝑘𝑘 Where d𝑉𝑉d𝑡𝑡 is the volume of liquid flowing through the tube per unit of time, ∆𝑃𝑃 pressure drop along the tube of length 𝑘𝑘 and 𝜂𝜂 is the viscosity of the liquid. The velocity at a distance 𝑟𝑟 from the center is given by: Figure 46. Sketch of 𝑟𝑟2 − 𝑟𝑟2 ∆𝑃𝑃 𝑟𝑟2c c − 𝑟𝑟2 d𝑉𝑉 8 𝜂𝜂 𝑟𝑟2 d𝑉𝑉 2 the tube. 𝑣𝑣(𝑟𝑟) = = = �1 − � (36) 4 𝜂𝜂 𝑘𝑘 4𝜂𝜂 d𝑡𝑡 𝜋𝜋𝑟𝑟4c 𝑟𝑟2c d𝑡𝑡 𝜋𝜋𝑟𝑟2c The volume flow also follows from the continuity equation: d𝑉𝑉 𝑟𝑟c = 2𝜋𝜋� 𝑟𝑟𝑣𝑣(𝑟𝑟) d𝑟𝑟 d𝑡𝑡 0 𝑟𝑟c 𝑟𝑟2 𝑟𝑟2 𝑟𝑟4 𝑟𝑟c 𝑟𝑟2 𝑟𝑟4 𝜋𝜋 = 2𝜋𝜋𝑣𝑣0 � 𝑟𝑟 �1 − 2�d𝑟𝑟 = 2𝜋𝜋𝑣𝑣0 � − 2� = 2𝜋𝜋𝑣𝑣 c c 0 � − 2� = 𝑣𝑣 𝑟𝑟 2 (37) 0 𝑟𝑟c 2 4𝑟𝑟c 0 2 4𝑟𝑟c 2 0 c with 𝑣𝑣0 = 𝑣𝑣(𝑟𝑟 = 0) = d𝑉𝑉 2 d𝑡𝑡 𝜋𝜋𝑟𝑟2c At 𝑟𝑟 = 𝑟𝑟c the shear rate is: d𝑣𝑣(𝑟𝑟 = 𝑟𝑟 ) 2𝑟𝑟 2 d𝑉𝑉 4 d𝑉𝑉 ?̇?𝛾 = c = − c2 2 = − (38) d𝑟𝑟 𝑟𝑟c 𝜋𝜋𝑟𝑟c d𝑡𝑡 𝜋𝜋𝑟𝑟3c d𝑡𝑡 124 HOW TO COAT THE INSIDE OF NARROW AND LONG TUBES WITH A SUPER-LIQUID-REPELLENT LAYER For the applied volume flow of d𝑉𝑉d𝑡𝑡 = 11 mL min -1 = 1.8 × 10-7 m3 s-1 and a tube radius of 𝑟𝑟c = 0.475 mm, the shear rate is: 4 d𝑉𝑉 |𝛾𝛾̇ | = 3 = 2 × 10 3 s-1 (39) 𝜋𝜋𝑟𝑟c d𝑡𝑡 BACTERIA E. COLI ADHESION TESTING AND IMAGING The tubes cut in half horizontally were fixed onto a sterile chambered glass with double- sided tape. Each chamber was covered with LB (Lysogeny broth)+Amp medium (1 mL, containing 100 µg mL-1 of Ampicillin as well as Arabinose added to a final concentration of 0.1%, in the following referred to as LB+Amp medium) containing E. coli expressing eGFP (108 cells mL-1 controlled through OD600, Stellar™ Competent, HST08, Clontech) and left for incubation at 37 ℃ for 7 days. The bacterial solution was refreshed daily to avoid nutrient depletion of the bacteria. The bacteria solution for the experiment was prepared by mixing 1 mL of the mother solution containing E. coli expressing eGFP with 4 mL of LB+Amp medium, followed by 2 h of incubation at 37 °C at 350 rpm. The optical density of the bacteria-solution was adjusted by dilution with LB+Amp medium. To prepare the samples for SEM investigations, after the 7-day incubation period, the culture medium was carefully removed, and the samples were washed with phosphate-buffered saline (PBS, 1 mL, 3 times). The phosphate-buffered saline was free of MgCl2 and CaCl2. Adhered bacteria were then fixated by adding a 2.5% (𝑉𝑉/𝑉𝑉) glutaraldehyde solution in PBS for 30 min at room temperature. The fixative was removed afterward, and the remaining ma- terial was washed off by thorough rinsing with buffer solution (PBS, 1 mL, 2 times). Sub- sequently, the adhered bacteria were dehydrated by successive ethanol soaking (i.e., soaking in water-ethanol mixtures, 25, 50, 60, 70, 80, 90, and 100% (𝑉𝑉/𝑉𝑉), 15 min each, and last step twice). Then the samples were placed in a vacuum dryer overnight at room tem- perature. Sterilized consumables were used throughout. To avoid charging, the dried sur- faces were sputter-coated with an nm-thick layer of Pt (BalTec MED 020 Modular High Vacuum Coating System, Argon at 2 × 10-2 mbar. 7 nm Pt and 30 mA). FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 125 DETECTION AND CALCULATION OF AVERAGE BACTERIA NUMBER ON BARE AND COATED PU TUBES Based on SEM images, the area covered by bacteria on the PU tubes was evaluated using the Analyze Particles plugin of ImageJ. Background subtraction was applied before thresholding to correct for an inhomogeneous background. The threshold was chosen us- ing the default method of ImageJ, being a modified IsoData algorithm. Statistics were de- rived from the evaluation of 14 to 24 images per sample (3 samples in total). The number of bacteria per image (and thus per area) was calculated by dividing the coverage area of the bacteria by the average area of a single bacterium. Afterward, the average number of bacteria per area of each sample was calculated. Finally, the average number of bacteria per area and corresponding standard deviation based on the samples was calculated. Figure 47 shows an example of this analysis, recognized bacteria areas are highlighted in cyan. Figure 47. SEM image of original data (left) and the recognized bacteria (right, cyan) on a bare PU tube. Due to the high surface roughness of the nanofilament-coated PU tubes, the bacteria could not reliably be detected using the above-mentioned procedure; therefore, bacteria per image (and thus area) were counted manually. Statistics were derived using at least 20 images per sample (2 samples for each nanofilament-coated PU tube). Afterward, the average number of bacteria per area, and the corresponding standard deviation over all samples was calculated. AUTHOR CONTRIBUTIONS F. Geyer fabricated the coated tubes, carried out the experiments unless stated otherwise below, and wrote the manuscript. M. D’Acunzi and M. Müller assisted with the coating of 126 HOW TO COAT THE INSIDE OF NARROW AND LONG TUBES WITH A SUPER-LIQUID-REPELLENT LAYER the tubes. C.-Y. Yang, P. Baumli, N. Encinas, and V. Mailänder carried out the bacteria incubation experiments. A. Kaltbeitzel contributed to the data analysis of the bacteria ex- periments. P. Baumli conducted the XPS measurements. F. Geyer, D. Vollmer, and H.-J. Butt contributed to experimental planning, data analysis, and manuscript preparation. All authors reviewed and approved the manuscript. ACKNOWLEDGMENTS This work was supported by the ERC Advanced Grant 340391 – SUPRO (H.-J.B.), the German Research Foundation (DFG) with the Collaborative Research Center 1194 (H.-J.B., F.G.), the European Union’s Horizon 2020 research and innovation program LubISS No. 722497 (D.V., P.B.), the DAAD-MoST sandwich scholarship (C-Y.Y.) and the Marie Skłodowska-Curie fellowship 660523-NoBios-ESR (N.E.). are acknowledged for technical support. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 127 6 POLYHEDRAL LIQUID MARBLES Florian Geyer,1 Yuta Asaumi,2 Doris Vollmer,1 Hans-Jürgen Butt,1 Yoshinobu Nakamura,3 and Syuji Fujii3 1Max Planck Institute for Polymer Research, Ackermannweg 10, 55128 Mainz, Germany 2Graduate School of Engineering, Osaka Institute of Technology, 5-16-1, Omiya, Asahi-ku, Osaka 535-8585, Japan 3Department of Applied Chemistry, Faculty of Engineering. Osaka Institute of Technology, 5-16-1, Omiya, Asahi-ku, Osaka 535-8585, Japan Abstract: A new type of armored droplet, a so-called polyhedral liquid marble, is intro- duced in this work. These armored liquid marbles consist of liquid droplets stabilized by hydrophobic hexagonal plates made of poly(ethylene terephthalate), which adsorb to the liquid-air interface. Depending on the specific combination of plate size and droplet di- ameter, the plates self-assemble into highly ordered hexagonally-arranged domains. Even tetrahedral-, pentahedral-, and cube-shaped liquid marbles composed of only 4 to 6 plates are demonstrated. During evaporation of the internal liquid, due to the high ad- sorption energy of the plates at the liquid-air interface, the overall surface area stayed constant, resulting in strongly deformed polyhedral liquid marbles. In line with this, highly asymmetric polyhedral liquid marbles and letters are obtained due to the strong interfacial jamming exerted by the rigid hexagonal plates. This is particularly pronounced for larger plate sizes leading to liquid marbles with unusually sharp edges (for example, rectangular edges). The polyhedral liquid marbles exhibit various stimuli-responsive be- haviors being exposed to water, ammonia, or tetrahydrofuran vapors. Air-driven polyhe- dral liquid marbles floating on water can reach velocities of several centimeters per sec- ond. Keywords: liquid marbles, interfacial jamming, wetting, stimuli-responsive This chapter is reprinted with permission from F. Geyer et al., Adv. Funct. Mater. 2019, 1808826. Copyright 2019, John Wiley and Sons Important Supporting Figures can be found in Appendix 10.2. 128 POLYHEDRAL LIQUID MARBLES INTRODUCTION Studying the adsorption of solid particles at air-water and oil-water interfaces is of broad interest.[181] Recently, various novel concepts and materials such as colloidosomes,[196, 257] armored bubbles,[182, 302] dry water,[179] and liquid marbles (LMs) were intro- duced.[184, 186-188, 191, 197] LMs are liquid droplets coated with particles, which are adsorbed at the liquid-gas interface. These armored droplets are non-wetting, soft, and elastic since the particles protect the inner liquid from wetting the substrate underneath.[303] LMs in air are different from emulsions and foams as they require particles for stabilization and can- not be stabilized by molecular surfactants. LMs have been utilized in various potential applications, such as transport of liquids,[200] miniature reactors,[194, 199] microcapsules,[192] micro bioreactors for blood analysis,[198] sensors,[190] accelerometers,[203] and pressure- sensitive adhesives.[193] LMs can be prepared using solid particles with various chemical compositions, including organic, inorganic, and composite particles. Most of the literature is concerned with (near) spherical particles and their aggregates. However, studies on LMs stabilized with non- spherical particles or stabilizers are rare. Thus, it is crucial to reveal and understand the relationship between solid particle shape and LM structure formation to utilize the LMs. Herein, we introduce a new type of LM, a so-called polyhedral LM. The polyhedral LMs are stabilized using hydrophobic hexagonal plates monodisperse in shape and size. The questions addressed in this study are: (1) Can hexagonal plates stabilize LMs? Which ratio between plate and droplet size can be achieved? (2) How do the plates self-assemble at a curved air-water interface, especially when the curvature and plate size are comparable? (3) What are the properties and functionalities of the polyhedral LMs? RESULTS AND DISCUSSION Poly(ethylene terephthalate) (PET) hexagonal plates (Figure 48a) with monomodal distri- butions in both size (Major axis: 𝐴𝐴 = 2𝑐𝑐 = 0.27 ± 0.04 mm (15% rel. error), 0.34 ± 0.01 mm (3% rel. error), 1.14 ± 0.04 mm (4% rel. error) and 2.48 ± 0.04 mm (2% rel. error); side length 𝑐𝑐 = 0.14 mm, 0.17 mm, 0.57 mm and 1.24 mm, respectively) and thickness (𝑑𝑑 = 40– 55 µm for all the samples) were used in this study (Figure S1 in Appendix 9.2 for optical and scanning electron microscopic (SEM) images). The original PET plates were hydro- philic and wetted by water, being unable to stabilize LMs. The plates were hydrophobized by modification with 1H,1H,2H,2H-perfluorodecyltrichlorosilane in n-hexane. Energy FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 129 dispersive X-ray (EDX) spectroscopy confirmed the surface coating of the PET plates by the silane (Figure S2). The static contact angle of a 10 µL water droplet measured on a fluorinated PET sheet was 110° ± 5° (critical surface tension was estimated to be approxi- mately 5 ± 3 mN m-1 by Zisman plot; the value is quite low and likely influenced by the roughness of the substrate; Figure S3 in Appendix 10.2) indicating a hydrophobic fluorinated surface. Water droplets (60 µL) were deposited onto the dried hydrophobic plates using a micropi- pette. During gentle rolling of the water droplets on the beds, the liquid was spontane- ously encapsulated by the plates, and LMs were formed (Figure 48b). The coating of the water droplets by the plates rendered them both hydrophobic and non-wetting. The plates adsorbed at the air-water interface of the droplet form a certain contact angle 𝜃𝜃 with re- spect to the side wall of the plates (Figure 48c,d), which is mainly determined by the hy- drophobic surface modification of the plates. Imaging of the lower side of a polyhedral LM using laser scanning confocal microscopy (LSCM) revealed that the contact line is pinned at the inner edge of the plates, roughly in parallel with the plate’s long surface and perpendicular to its normal direction (𝜃𝜃 ≈ 90°). The air-water interface and thus the drop- let (Figure 48d, bottom; shown in blue) was trapped at the inner side of the plates. In accordance with that, the distance h between the outer side of the plates (Figure 48d, bot- tom, reflection on the outer side of the plate shown in yellow) and the air-water interface was ≈40 µm, which corresponds to the thickness of the plates. This resembles the situation of a water droplet on a superhydrophobic surface consisting of fluorinated micropil- lars.[61-62] Hence, the hexagonal plates act as a mobile micropillar layer. The LMs remained stable even after transfer from the plate beds onto solid substrates, such as PET or hydro- philic glass slides. Optical microscopy images show the microstructure of the LMs prepared using 0.27 mm, 0.34 mm, 1.14 mm, and 2.48 mm-sized plates, placed on a hydrophilic glass slide (Figure 48e, left to right). Polyhedral LMs were formed in all cases. Plates spontaneously arranged in a hexagonal two-dimensional crystal-like structure, which is readily apparent on the LM surfaces of the 0.34 mm and 1.14 mm-sized plate systems (Figure 48f). If the droplet diameter (here: 𝐷𝐷 ≈ 6 mm) is considerably larger than the plate size, the droplet curvature is small compared to the plate size, and the plates can fully attach to the droplet, forming highly ordered two-dimensional domains on the surface of the droplet. LMs pre- pared with the 0.34 mm-sized plates showed the highest order. The plates spontaneously 130 POLYHEDRAL LIQUID MARBLES self-assembled to form well-ordered arrays on the droplet’s surface just by rolling, even though the LMs were formed from randomly oriented plate beds. Figure 48. a) Schematic of the hexagonal fluori- nated PET plates used to stabilize polyhedral liquid marbles (LMs) with thick- ness d, main axis A, and side length s. b) Sche- matic of a polyhedral LM. c) Optical microscopic im- age of plates (yellowish; with A = 2.48 mm) ad- sorbed on a 60 µL water droplet (with Brilliant Blue dye, 20 mg L-1). Scale bar: 0.5 mm. d) The fluori- nated plates adsorb to the air-water interface form- ing a contact angle θ. La- ser scanning confocal mi- croscopy (LSCM) image from underneath a poly- hedral LM (15 µL water containing ATTO488, fluo- rescence shown in blue, reflections in yellow) shows that θ ≈ 90°. The distance h between the outer side of the plate and the air-water interface is ≈40 µm. e) Optical micros- copy images of polyhe- dral LMs prepared with different plate sizes (A = 0.27 mm, 0.34 mm, 1.14 mm, and 2.48 mm; left to right) and different droplet volumes (60 µL, 30 µL, and 15 µL; top to bottom). Scale bar: 1 mm. f) Magnified microscopic images revealing the mi- crostructure of the polyhedral LMs with different plate sizes (A = 0.27 mm, 0.34 mm, 1.14 mm, and 2.48 mm; left to right). Scale bars: 0.5 mm. g) By tuning the plate to water ratio (15 µL droplets with bril- liant blue dye) using the largest 2.48 mm-sized plates (a total of 4 to 6 plates from left to right), different polyhedral morphologies were obtained. Scale bar: 1 mm. h) SEM image of a dried cyanoacrylate-treated 3 µL polyhedral LM using 0.34 mm-sized plates. Scale bar: 200 µm. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 131 Parts of the plates stuck out of the LM’s surface into the air phase in the case of the largest 2.48 mm-sized plate system, where the droplet diameter (𝐷𝐷 ≈ 6 mm) was comparable to the plate size. The plates were tangent to the LM surface. A significantly lower degree of order was observed for the smallest 0.27 mm-sized plate system correlating to its higher polydispersity[304] and its smaller size.[305] The diameters of the LMs were varied in a range between approximately 6 mm and 2.5 mm simply by changing the volume of the inner water droplet (60 µL–15 µL) (Figure 48e; top to bottom). For the 0.27 mm and 0.34 mm-sized plate systems, the de- crease of droplet volume yielded more spherical LMs. Since the plate sizes were much smaller than the radii of the droplets, the plates could fully attach to the droplets’ surfaces. The 1.14 mm-sized plates started to stick out of the LM surfaces into the air phase for droplet volumes ≤30 µL. The droplet curvature increased, and the stiff plates could only partially attach to the droplet surface. Using the 2.48 mm-sized plates, polyhedral LMs with a near-spherical shape formed for 30 µL droplets. Notably, when reducing the drop- let volume to 15 µL (𝐷𝐷 ≈ 2.5 mm), LMs with various polyhedral shapes, including tetrahe- dron, pentahedron, and cube, were obtained by tuning the plate to water ratio (Figure 48g). Depending on the number of plates adsorbed on the water droplet, the polyhedral LM spontaneously assumed the different shapes. Water droplets could be sta- bilized using as few as 4 to 6 plates. Due to the shape and rigidity of the hexagonal plates, the droplet could not be fully covered by the plates, leading to LMs with open faces. This is especially apparent for the polyhedral LM with 5 plates (Figure 48g; middle). Here, the hexagonal plate on the top and bottom of the LM kept the axial plates apart, whereas the capillary forces kept them in place, yielding a very uniquely shaped LM. Large LMs using 0.34 mm and 2.48 mm-sized plates containing 120, 480, and 960 µL water drops were fabricated. Their shapes changed from oblong to puddle with an increase of drop volume (Figure S4 in Appendix 10.2). Crystalline two-dimensional array structures developed in large areas. Despite the difference in the nature of the materials and length scales, the patterns formed on the LM surfaces are similar to the honeycomb structure built by honey bees in their nests. For the large plates, the size of the crystalline domains continually increased as the droplet/plate size ratio became larger. On the other hand, using the smaller 0.34 mm-sized plates, crystalline domain sizes hardly changed for drop volumes of 480 µL and 960 µL. Up to 50 plates per two-dimensional crystal were observed, corresponding to domain sizes of approximately 3–4 mm. 132 POLYHEDRAL LIQUID MARBLES The liquid-air interface of a 3 µL polyhedral LMs was solidified by exposing it to ethyl 2-cyanoacrylate vapor, the main component of superglues. Ethyl 2-cyanoacrylate readily absorbs and polymerizes at air-water interfaces,[306] solidifying the water interface be- tween and underneath the plates. After superglue treatment, the polyhedral LM pre- served its spherical, polyhedral morphology and did not collapse during/after the drying process, indicating that the surface of the LM was rigid (Figure 48h). The hexagonally- packed plate arrays were mostly preserved. Figure 49. a,b) Microscopic images from the top (a) and side view (b) showing morphological changes of polyhedral LMs (15 µL; 0.34 mm-sized plates) during evaporation. Scale bars: 0.5 mm. c) Comparison of the evaporation of a polyhedral LM (15 µL; 0.34 mm-sized plates) at 36 ± 5% relative humidity (RH) versus a 15 µL water droplet on a superhydrophobic substrate (average of three repeats at an avg. RH of 37 ± 5%) at 21–22 °C. d) Weight loss during evaporation of polyhedral LMs (15 µL; 0.34 mm-sized plates) at different humidities (56 ± 5% RH, 45 ± 5% RH, and 36 ± 5% RH). Detailed SEM images show that only a monolayer of the hexagonal plates was adsorbed at the air-water interface and confirm 𝜃𝜃 ≈ 90° obtained by the LSCM images (Figure S6 in Appendix 10.2). FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 133 When drying the polyhedral LMs, the morphology changed. As an example, drying of polyhedral LMs of 15 µL volume armored with 0.34 mm-sized plates is shown in Figure 49a,b (Figure S7 for examples of the 0.27 mm and 1.14 mm-sized plate systems). Slow evaporation of the inner water of the LM induced buckling on the LM’s surface. This can be explained by the surface jamming effect of the plates at the air-water interface due to the high adsorption energy 𝛥𝛥𝐺𝐺 at the interface. The adsorption energy (𝛥𝛥𝐺𝐺)[184, 307] is de- fined as the difference in free energy between a system where one face of a plate is in contact with water as compared to a system where the plate is fully in contact with air plus a free air-water interface. Thus, 𝛥𝛥𝐺𝐺 can be estimated using Eq. 40 (see Experimental Sec- tion 6.4.14 for details). ∆𝐺𝐺 = 3√3𝑐𝑐2(𝛾𝛾AS + 𝛾𝛾AW − 𝛾𝛾WS) (40) 2 where 𝛾𝛾 = 5 mN m-1AS , 𝛾𝛾WS = 30 mN m-1, and 𝛾𝛾AW = 73 mN m-1 are the air-plate, water- plate, and air-water interfacial energies, respectively (𝛾𝛾𝑊𝑊𝑆𝑆 was determined using Young’s equation with a water contact angle of 110˚). Using Eq. 40, the ∆𝐺𝐺 values of the plates ad- sorbed at air-water interfaces were calculated to be 2 × 10-9 J, 4 × 10-9 J, 4 × 10-8 J, and 2 × 10-7 J for 0.27 mm, 0.34 mm, 1.14 mm, and 2.48 mm-sized plates, respectively. These values are more than three orders of magnitude higher than those of micrometer-sized spherical particles adsorbed at air-water and oil-water interfaces.[183] The plates did not detach or flip off during evaporation, because of the high adsorption energy. The total surface area of the LM stayed almost constant, and the loss of volume was compensated by deformation. For comparison, a bare water droplet placed on a su- perhydrophobic substrate decreased its volume via evaporation while remaining nearly spherical (Figure S8). In contrast to the LM, the air-water interfacial area decreased along with the decrease of droplet volume. The water evaporation of the LMs and bare water droplets (Figure 49c,d) was evaluated gravimetrically at 21–22 °C and three humidities varying between 35% and 55% relative humidity (RH). Surprisingly, the mass and thus volume decreased almost linearly with time for the LMs. The water evaporation rate of a polyhedral LM of 15 µL volume was ≈0.14 mg min-1, and the water was completely evaporated after ≈105 min at 36 ± 5% RH. At 45 ± 5% and 56 ± 5% RH, the evaporation of the inner liquid slowed down, and evapo- ration times increased to ≈125 min and ≈155 min, respectively. The evaporation of bare water droplets of the same volume followed the 𝑉𝑉2/3 law,[308] implying that evaporation is determined by the shrinking spherical shape of the droplet (Figure S9). At 21–22 °C and 134 POLYHEDRAL LIQUID MARBLES avg. RH of 37 ± 5%, the droplet was evaporated after ≈110 min. Thus, even though the polyhedral LM was armored with hexagonal plates, the evaporation of the inner liquid was slightly faster than the evaporation of a bare droplet since the LM retained its surface areas whereas the bare droplet decreased its surface area.[309] The final stage of drying is interesting: the LMs broke down due to gravity and left cohesionless residues of plates after complete drying. Just before complete evaporation, the plates stuck to each other due to lateral capillary forces mediated by residual water. Spherocylindrical polyhedral LMs (Figure 50a) with aspect ratios larger than 10 can be created by either (1) merging polyhedral LMs, by (2) withdrawing liquid from the marble, or by (3) placing excess plates on prepared LMs and deforming those by external forces. To obtain a cylindrical shape, the LMs were then rolled back and forth along one axis. Being able to create cylindrical LMs is strikingly different from bare droplets, where the surface area is minimized in order to minimize the interfacial energy. The LMs behaved in an inelastic manner, and the long cylindrical shapes did not undergo a Rayleigh-Plat- eau capillary instability. The inelastic characteristics arise from the steric jamming of plates at the air-water interface. Jamming of colloidal particles at interfaces has already been employed to kinetically trap fluid morphologies, including non-spherical armored bubbles and droplets[310-312] as well as bicontinuous interfacially jammed emulsion gels (bijels).[313] The shapes of the jammed LMs can be plastically rearranged to various stable configurations if subject to inhomogeneous external stresses (Figure 50b,c). The high as- pect ratio cylindrical LM could be rearranged to write the letter “L” (Figure 50a to Figure 50b) by applying an external force using a spatula. Additionally, a dumbbell-like shape could be obtained (Figure 50c). We found that the large plates exhibit an even stronger interfacial jamming. Deposition and rolling of water droplets on the plates al- ready yielded highly deformable polyhedral LMs. Thus, the liquid letter “A” was easily formed using the 2.48 mm-sized plates (Figure 50d). This principle was expanded to trans- parent 𝐴𝐴 = 2.35 ± 0.03 mm-sized plates (Figure S10). By merging a few LMs, the initials of polyhedral liquid marble “PHLM” were written (Figure 50e, Movie S1 shows the strong interfacial jamming behavior and formation of letters within a few seconds). It should be noted that these LMs can feature extremely sharp edges having an angle of 90° (see inset in Figure 50e). FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 135 Figure 50. a) Spherocylindrical jammed polyhedral marble with an aspect ratio of 12 prepared with 0.34 mm-sized plates and containing water. Scale bars: 5 mm (left), 2 mm (right). b) Polyhedral letter “L” by deforming the high aspect ratio LM. Scale bar: 5 mm. c) Dumbbell-like shaped polyhedral LM with con- vex and concave areas. Scale bar: 5 mm. d) Polyhedral letter “A” showing strong interfacial jamming and shape control, formed with 2.48 mm-sized plates. Scale bar: 5 mm. e) Initials of the term polyhedral liquid marble: PHLM, formed using transparent 2.35 mm-sized plates and dyed water. The edges of the letters can be very sharp (see inset of letter “L”). Scale bar: 5 mm, 0.5 mm (inset). f) The edge angle δ of the polyhedral LMs can be tuned between >120° and <45°. Scale bar: 0.5 mm. g) Merging of two polyhedral LMs having a blue and a pink dye to form the letter “Q.” The mixing of the two dyes can be observed inside of the resulting LM. Scale bar: 5 mm. We tuned the edge angle 𝛿𝛿 of a polyhedral LM between >120° and <45° (Figure 50f). Such well-defined edges were previously not available with common LMs prepared using spherical particles or particle aggregates. By merging two LMs containing different dyes (pink and blue dye), the letter “Q” was formed (Figure 50g). The mixing of the dyes inside of the resultant LM could be observed through the transparent PET plates, indicating that the liquid phases became continuous. To show that the jammed interface is not a thermodynamically stable state, we exposed a cylindrical polyhedral LM containing hygroscopic saturated LiCl solution to water vapor (Figure S11a in Appendix 10.2). The cylindrical shape gradually changed to a spherical shape during water uptake from the vapor phase (due to an increase of droplet volume 136 POLYHEDRAL LIQUID MARBLES and surface area), indicating relaxation of the jammed interface. The polyhedral LM grown via water absorption was now placed under dry atmosphere in the presence of sil- ica gel (Figure S11b in Appendix 10.2). The LM size decreased, keeping its spherical shape and eventually started buckling. After 50 h, the LM was dried. This dried polyhedral LM was now exposed to humid atmosphere again, leading to the growth of the LM. It resumed to its previous spherical shape (Figure S11c in Appendix 10.2), yet the LM could not return to its non-equilibrium jammed cylindrical shape. This further demonstrates that the jam- ming can be released by an external stimulus. Polyhedral LMs are simultaneously responsive to different stimuli (Figure 51a,b). A trans- parent polyhedral LM (prepared with 𝐴𝐴 = 1.13 ± 0.04 mm-sized transparent PET plates) containing 30 µL water with the pH-responsive dye phenolphthalein was first exposed to ammonia vapor (Figure 51c). The absorption of ammonia increases the pH from neutral to slightly basic, resulting in a color change of the liquid inside of the LM at pH = 8.2. Close observation of the LM (Figure S12 in Appendix 10.2) shows that the ammonia vapor penetrated through the gaps between the PET plates and slowly diffused and distributed over the whole liquid marble. After complete color change, the LM was exposed to tetra- hydrofuran (THF) vapor. Within the first 30 s of exposure, the volume of the LM increased slightly. Then, suddenly after 33 s, the polyhedral LM collapsed. Due to the uptake of THF, the surface tension of the liquid phase inside of the LM decreased. At a certain point, the liquid started wetting the plates, most likely inducing a Cassie-to-Wenzel-like[62] transition. The liquid started contacting and spreading on the glass slide underneath. A photograph, captured from underneath a broken LM (Figure S13 in Appendix 10.2) prepared with non- transparent plates and exposed to THF, shows that the liquid spread toward a certain di- rection (here: toward the left). Surprisingly, although polyhedral LMs exhibit only one monolayer of plates, they can be transferred and float on a planar water surface (Figure 51d). After transfer, the liquid-air interfaces of the water inside and underneath the LM are separated by just a monolayer of plates (Figure S14 in Appendix 10.2). The polyhedral LM remained stable on the water surface and could float for more than 1 h. The distance between the two interfaces ℎ should be close to the plate thickness 𝑑𝑑 ≈ 40 µm. Nonetheless, this tiny air gap is enough to separate the two water interfaces and let the polyhedral LM float. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 137 Figure 51. a,b) Schematic of a stimuli-responsive polyhedral LM. Due to the absorption of (a) NH3 vapor and the resulting pH changes, the LM (containing phenolphthalein) changes its color from colorless and transparent to pink. The absorption of (b) tetrahydrofuran (THF) leads to a decrease of surface tension of the inner liquid and thus a Cassie-to-Wenzel-like transition between the hexagonal plates. c) Stimuli- responsive polyhedral LM (30 µL, 1.13 mm-sized transparent plates) exposed to NH3 vapor for 120 s and subsequently to THF vapor. Scale bar: 2 mm. d) A polyhedral LM (15 µL; 0.34 mm-sized plates) floating on a planar water surface. The two air-water interfaces are separated by just a monolayer of plates, which corresponds to a distance of h ≈ d ≈ 40 µm. Scale bar: 1 mm. e) Locomotion of a 15 µL polyhedral LM on a water surface driven by air blows. Scale bar: 5 mm. f,g) Path and corresponding velocity analysis of air- driven polyhedral LM locomotion on water for 2 blows. h) Stimuli-induced disruption of a 15 µL polyhedral LM by the addition of THF to the water surface. The LM contained a 3 wt% NaOH solution, and the water reservoir contained phenolphthalein dye. Scale bar: 1 mm. Fast locomotion of the LM on water can be induced by air blowing (Figure 51e). Depend- ing on the strength of the blow, the LMs traveled several centimeters and reached peak velocities of almost 10 cm s-1. This results in accelerations of up to 60 cm s-2 and forces of up to 10 µN (Figure 51f,g). A floating LM was disrupted by addition of THF to the water 138 POLYHEDRAL LIQUID MARBLES surface (Figure 51h, Movie S2 shows the locomotion of a polyhedral LM induced by air blowing and disruption via THF). The LM contained 15 µL of a highly basic 3 wt% NaOH solution. After disruption, the LM changed the pH and color of the underlying water res- ervoir, which contained phenolphthalein. Thus, the LMs may be used as pH and surface tension sensitive sensors. CONCLUSION In conclusion, LMs stabilized by hydrophobic hexagonal PET plates are introduced. The arrangement of the plates depends on the ratio of the plate diameter 𝐴𝐴 and the droplet diameter 𝐷𝐷. For small droplets, where 𝐷𝐷 and 𝐴𝐴 are of the same order of magnitude, the self-assembly of the hexagonal plates on the surface of the LMs results in the first exam- ples of uniquely shaped polyhedral LMs, including cubic, tetrahedral and pentahedral structures. A similar yet different phenomenon is known from the so-called capillary ori- gami where a highly elastic sheet wraps a droplet.[314] In contrast, the PET plates are rigid and not bend by the capillary forces. For large droplets (𝐷𝐷 ≫ 𝐴𝐴), the surfaces of the LMs are composed of two-dimensional polycrystalline arrays of plates. The high adsorption energy of the plates at the air-water interface of the LMs prevents the plates from detach- ing, e.g., during evaporation, and leads to interfacial jamming. Highly jammed non-equi- librium shaped LMs, such as cylindrical droplets and letter-shaped marbles, can be formed. Those LMs exhibit tunable and extraordinarily sharp edges. The polyhedral LMs are stable enough to move on planar water surfaces and can be used as sensors as they are responsive to different stimuli. The use of the hexagonal plates enables unprece- dented shape control, and the methodology for generation of these crystalline structures on curved surfaces is simple, scalable, and based on sustainable materials. EXPERIMENTAL SECTION MATERIALS Experiments were carried out at room temperature. The following chemicals were used: 1H,1H,2H,2H-perfluorodecyltrichlorosilane (PFDTS, 96%, TCI), n-hexane (96%, Wako), ethanol (95%, Sigma-Aldrich), tetrahydrofuran (SAJ first grade, Sigma-Aldrich), ammonia solution (28 wt%, Sigma-Aldrich), lithium chloride (99.0%, Wako), sodium hydroxide (≥97%, Sigma-Aldrich), ethyl 2-cyanoacrylate (Aron Alpha Extra Sokkotayoto, Toagosei FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 139 Co., Ltd.), Brilliant Blue (San-Ei Gen F.F.I., Inc.), Food Red No. 106 (San-Ei Gen F.F.I., Inc.), phenolphthalein (extra pure, Kanto Chemical Co., Inc.), ATTO488 (ATTO-TEC). Purified and deionized water from Advantec MFS RFD240NA (GA25A-0715) and from Sartorius Arium Pro VF was used for all experiments. Hydrophilic PET plates (A = 0.34 ± 0.01 mm) were purchased from Matsumura Kogei Co., Ltd. The other hydrophilic PET plates (A = 0.27 ± 0.04 mm, 1.14 ± 0.04 mm, and 2.48 ± 0.04 mm) were purchased from Matsuno Industry Co., Ltd. Transparent hydrophilic PET plates (A = 1.13 ± 0.04 mm and 2.35 ± 0.03 mm) were obtained from Nakajima Metal Leaf, Powder Co. Ltd. Reagents were used as received. HYDROPHOBIZATION OF THE PLATES The pristine plates (5–10 g) were washed repeatedly with ethanol (200–400 mL) until the supernatant remained clear (3–7 repeats). The washed plates were dried in vacuum over- night. Afterward, the dried plates were modified with PFDTS. Therefore, 3–10 g of the plates was suspended in a solution of 45 µL PFDTS in 45 mL n-hexane, and the mixture was shaken for approximately 30 s. During the reaction time of 30 min, the sedimented suspension was shaken every 5 min to ensure a homogeneous coating of the plates. Sub- sequently, the modified plates were rinsed at least twice with n-hexane. PREPARATION OF POLYHEDRAL LIQUID MARBLES The polyhedral liquid marbles (LMs) were prepared by simply rolling water droplets over a plate bed on a glass substrate or rough poly(chlorotrifluoroethylene) (PCTFE) dish. In the case of the tetrahedral-, pentahedral-, and cube-shaped LMs, 15 µL droplets were rolled over individual 2.48 mm plates placed on a surface-roughened PCTFE Petri dish. This allowed the droplet to pick up individual plates. Afterward, the prepared LMs may be transferred to other substrates like glass slides using a (plastic) spoon. PREPARATION OF JAMMED POLYHEDRAL LIQUID MARBLES To obtain jammed polyhedral LMs with non-equilibrium shapes, LMs prepared according to the previous paragraph (on roughened PCTFE Petri dish) were either merged, or water was withdrawn from the LM (half of the initial volume). Alternatively, an LM was pre- pared, and excess amounts of plates are placed on top of the LM. Now the LM could be deformed using a spoon, for instance, retaining the deformed shape. To prepare complex structures, the techniques mentioned above are combined to obtain the final result. LMs prepared with the large 2.48 mm and 2.35 mm-sized plates showed such strong interfacial 140 POLYHEDRAL LIQUID MARBLES that they could be easily deformed even without the steps mentioned above (see Movie S1). Photographs were taken using an optical microscope and digital cameras. The colored LMs were dyed using Brilliant Blue (20 mg L-1) and Food Red No. 106 (appears pink, 20 mg L-1). POLYMERIZATION OF ETHYL 2-CYANOACRYLATE ON THE AIR-WATER INTERFACE OF POLY- HEDRAL LMS A polyhedral LM (3 µL, 0.34 mm-sized plates) was placed on a pedestal in a reaction cham- ber with a vessel containing ethyl 2-cyanoacrylate. The ethyl 2-cyanoacrylate was then heated at 50 °C to saturate the gas phase with its vapor. Now, the ethyl 2-cyanoacrylate absorbed and polymerized at the air-water interface, solidifying the interface.[306] After a reaction time of 30 min, the polymerization was finished. Subsequently, the LM was dried up and imaged using the scanning electron microscope (SEM). HYGROSCOPIC POLYHEDRAL LIQUID MARBLES A polyhedral LM containing ≈100 µL of a saturated aqueous LiCl solution was prepared using 0.34 mm-sized plates and placed on a PCTFE Petri dish. A small watch glass (5 cm diameter) containing 500 µL water was placed near the LM, and the PCTFE dish was closed by placing another glass Petri dish onto it. For further sealing, Parafilm was applied. Af- terward, the water uptake was observed using an optical microscope for 4 h. After that, the water reservoir was replaced by silica gel, and the drying process was observed for 50 h. Finally, the silica gel was replaced by a water reservoir again, and the water uptake was observed for 4 h. STIMULI-RESPONSIVE POLYHEDRAL LIQUID MARBLES A polyhedral LM (30 µL aqueous phenolphthalein solution, 1.1 mm-sized transparent plates) was placed on a large glass slide. The phenolphthalein solution was prepared by mixing 10 mL water with 10 µL 5 wt% ethanolic phenolphthalein solution. A small watch glass (5 cm diameter) containing 200 µL ammonia solution (28 wt%) was placed next to the LM, and both were covered by a glass Petri dish. The color change of the LM was ob- served using the optical microscope. After the color change was completed, the watch glass containing ammonia solution was removed and replaced by a watch glass containing 400 µL tetrahydrofuran (THF). Absorption of THF and breakage of the LM was observed using the optical microscope. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 141 STIMULI-RESPONSIVE POLYHEDRAL LIQUID MARBLES ON WATER A polyhedral LM (15 µL aqueous 3 wt% sodium hydroxide solution, 0.34 mm-sized plates) was transferred to a water surface (50 mL water containing 45 µL 5 wt% ethanolic phenol- phthalein solution, in a glass Petri dish with a diameter of 10 cm). The LM was moved on the water surface by using an air blower. Afterward, a total of 330 µL THF was added to the water surface in 5 steps. Digital video cameras (Sony Handycam HDR-CX270 V with 30× optical zoom lens and Sony Alpha 7 II with Sony SEL2470Z Zeiss objective) were used to record photographs and movies. The velocity, acceleration, and applied force of and on the LMs were analyzed using the Keyence VW-9000 MotionAnalyzer software. The po- sition of the center of mass 𝑟𝑟(𝑡𝑡) of the LMs was determined in each frame. The displace- ment Δ𝑟𝑟(𝑡𝑡) and the velocity 𝑣𝑣(𝑡𝑡) = Δ𝑟𝑟(𝑡𝑡), with Δ𝑡𝑡 being the video rate (30-1 s), were ob- Δ𝑡𝑡 tained from the time course of 𝑟𝑟(𝑡𝑡). The acceleration 𝑎𝑎(𝑡𝑡) = Δ𝑣𝑣(𝑡𝑡), and the force 𝐹𝐹(𝑡𝑡) = Δ𝑡𝑡 𝑚𝑚𝑎𝑎(𝑡𝑡), with 𝑚𝑚 being the mass of the LM, were obtained from the changes of velocity Δ𝑣𝑣(𝑡𝑡). Thus, we could evaluate the velocity, acceleration, and force of the polyhedral LMs on the water surface, as previously reported.[315-316] PREPARATION OF THE SUPERHYDROPHOBIC SURFACES Candle soot-based superhydrophobic (and superoleophobic) surfaces were prepared ac- cording to ref.[41] The surfaces typically show static contact angles of >160° and 155° ± 5° as well as roll-off angles of 1° ± 1° and 5° ± 2° for 6 µL water and hexadecane droplets, respec- tively. The receding contact angle of water droplets was >160° as well. The PCTFE Petri dishes were roughened using sandpaper (grit size P200–P600).[317] Afterward, 15 µL drop- lets showed an almost spherical shape and could easily roll on the roughened Petri dish when tilted by 10–20°. OPTICAL MICROSCOPY Polyhedral LMs were observed using an optical microscope with an attached CCD camera (3.0 Megapixels USB Camera GR300BCM2, Shodensha Co., Japan). If not stated otherwise, the LMs were prepared and placed on a hydrophilic glass slide or a PCTFE dish and ob- served under the microscope under various magnifications. The scale bars were obtained by using a micrometer scale ruler. Brightness, contrast, and color temperature were ad- justed to match between the individual images. For some images, background subtraction using ImageJ was applied. 142 POLYHEDRAL LIQUID MARBLES LASER SCANNING CONFOCAL MICROSCOPY Laser scanning confocal microscopy (LSCM) images were taken using an inverted confo- cal microscope (Leica TCS SP8 SMD, HC PL APO 40×/1.10 CORR CS2 water objective) using Leica LAS X software. A polyhedral LM (15 µL water containing ATTO488 dye, 0.34 mm- sized plates) was placed onto a glass slide (≈150 µm thickness), and the lower side of the LM was imaged using the LSCM. The fluorescence of ATTO488 in water is shown in blue, and the reflection of light (at the lower side of the plate) is shown in yellow (Figure 48d). Reflection and fluorescence signals were recorded simultaneously. It should be noted that the water phase behind the hexagonal plate is not visible because the used plate was non- transparent. Only a strong reflection at the lower side of the plate is visible. SCANNING ELECTRON MICROSCOPY Scanning electron microscopy (SEM) images were taken with a VE-8800 instrument (Keyence, Japan) at gun voltages of 12 kV. To avoid charging, the samples were sputter- coated with gold using an Au coater (SC-701 Quick Coater, Elionix, Japan). ADSORPTION ENERGY OF A PLATE The adsorption energy (𝛥𝛥𝐺𝐺)[184, 307] is the free energy required for a plate to desorb from the air-water interface into the air phase and is defined as the difference in free energy between a system 𝐸𝐸Air, where the plate is entirely in contact with air plus a free air-water interface, as compared to a system 𝐸𝐸Interface, where one face of a plate is in contact with water, and the other faces of the plate are in contact with air (𝐸𝐸Air − 𝐸𝐸Interface). The energies 𝐸𝐸Air and 𝐸𝐸Interface for hexagonal plates are given by Eqs. 41 and 42, respectively. 𝐸𝐸 2 3√3 2Air = �3√3𝑐𝑐 + 6𝑐𝑐𝑑𝑑�𝛾𝛾AS + 𝑐𝑐 𝛾𝛾AW (41) 2 𝐸𝐸 3√3Interface = � 𝑐𝑐2 + 6𝑐𝑐𝑑𝑑� 𝛾𝛾 3√3 2 2 AS + 𝑐𝑐 𝛾𝛾 (42) 2 WS where s is the side length of the hexagonal plate; d is the thickness of the plate; and 𝛾𝛾AS, 𝛾𝛾WS , and 𝛾𝛾AW are the air-plate, water-plate, and air-water interfacial energies, respec- tively. Using Eqs. 41 and 42, 𝛥𝛥𝐺𝐺 is given by Eq. 43. ∆𝐺𝐺 = 3√3𝑐𝑐2(𝛾𝛾 2 AS + 𝛾𝛾AW − 𝛾𝛾WS) (43) FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 143 Here, it was assumed that only one large plane of the plate is in contact with water, and side planes remained in contact with air. This simple approximation neglects the fact that the droplet has a spherical shape while the plate-water interface is planar. AUTHOR CONTRIBUTIONS F. Geyer fabricated the polyhedral liquid marbles and carried out the experiments. Y. Asaumi assisted with the preparation of the liquid marbles and took the SEM images. Y. Asaumi conducted the locomotion analysis of the air-driven liquid marbles on water. S. Fujii and F. Geyer wrote the manuscript. S. Fujii suggested the project; S. Fujii and F. Geyer planned the experiments. All authors reviewed, edited, and approved the man- uscript. ACKNOWLEDGMENTS We thank for technical support. is thanked for the dyes (Brilliant Blue and Food Red No. 106). This work was supported by Grant-in-Aid for Scientific Research (B) (JSPS KAKENHI Grant Number JP16H04207; S.F.), Scientific Research on Innovative Areas “Engineering Neo- Biomimetics (No. 4402)” (JSPS KAKENHI Grant Numbers JP15H01602 and JP25120511; S.F.), “New Polymeric Materials Based on Element-Blocks (No. 2401)” (JSPS KAKENHI Grant Numbers JP15H00767 and JP25102542; S.F.), “Molecular Soft Interface Science (No. 2005)” (JSPS KAKENHI Grant Number 23106720; S.F.), ERC Advanced Grant No. 340391 “SUPRO” (H.-J.B., F.G.), the German Research Foundation (DFG) with the Collaborative Research Center 1194 (H.-J.B., F.G.), and the European Union’s Horizon 2020 research and innovation program LubISS No. 722497 (D.V.). 144 IMPACT AND OUTLOOK 7 IMPACT AND OUTLOOK This thesis contributes to the fabrication, application, and understanding of superhydro- phobic and super-liquid-repellent surfaces as well as the basic understanding of drop movement on such surfaces. To fabricate super-liquid-repellent surfaces, silicone nano- filament coatings were employed. These coatings can be applied to a broad range of dif- ferent surfaces, including flexible and stiff, organic and inorganic substrates. By applying nanofilament coatings to such substrates, various kinds of applications can be realized, ranging from self-cleaning surfaces, antibacterial surfaces to gas exchange membranes. Self-cleaning surfaces are important for various environments, e.g., for walls, windows, or as cover for solar cells. A key parameter to maintain the integrity of superhydrophobic surfaces and their self-cleaning properties is to prevent clogging of the pores. Therefore, the pores need to be smaller than the contaminant size. Even nanopar- ticles with a diameter of 600 nm can be cleaned off a nanoporous nanofilament surface. However, not only the particle size is important but also the polarity. While hydrophobic particles only cause minor harm to such surfaces, small hydrophilic particles dispersed in a liquid are detrimental. A pore size of a few hundred nanometers (<500 nm) is suitable to keep off most possible contaminations. In industrial environments with higher nano- particle concentrations, the pore size may be reduced accordingly. The self-cleaning properties of the coating are also reflected in the low lateral adhesion and friction forces of drops on the coating. As soon as self-cleaning fails, the adhesion increases. The DAFI method used to quantify the lateral adhesion and friction forces is a highly promising tool to characterize surfaces and self-cleaning and can be applied to various substrates. Fur- thermore, the method may be used to investigate the influence of other contamination types, such as biological contaminations. Superomniphobic membranes based on nanofilament-coated fabrics can be utilized as gas contactor membranes. Even highly concentrated amine solutions can be repelled, yielding enhanced CO2 exchange performances compared to conventional membranes. Wetting of the membrane is prevented. In times when humanity is threatened by global warming, and CO2 emissions reach new record highs, research on such topics is highly relevant and may contribute to the search for efficient methods for carbon capture and storage. The superomniphobic membranes are not restricted to CO2 capture and may also be useful for the separation and capture of other gases, such as H2S and SO ,[318-319]2 or re- lated applications like membrane distillation.[320] Recently, new liquids for CO2 capture, FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 145 such as ionic liquids showing higher loading capacities, were introduced.[321-323] Here, a superomniphobic membrane is essential since conventional membranes cannot repel such low surface tension liquids. Factors like impalement pressure and long-term stabil- ity of the Cassie state need to be carefully investigated and related to structural features of such membranes. Finally, the results may be transferred to porous hollow fiber membranes to maximize the packing density and thus the exchange efficiency per volume. Here, nanofilament coatings are a suitable candidate to coat such fibers and tubes. Meter-long and narrow tubes can be coated with nanofilaments after activation, e.g., by Fenton reaction. The coating can be applied to a wide range of materials. The coatings also show antibacterial properties and may be suitable for medical devices like catheters to reduce infections of patients due to restrained biofilm formation. A major benefit of these coatings is that they maintain these properties independent of the wetting state, i.e., even in a fully wetted superhydrophilic state. These properties are also important for other applications where the surfaces are in contact with aqueous solutions, such as water pipelines or membrane distillation, to prevent fouling of such systems.[324] Liquid marbles and self-cleaning of superhydrophobic surfaces are closely related. Liquid marbles can be formed during the pickup of contamination particles. During self-clean- ing, the particles accumulate at the liquid-air interface at the bottom of the drop and are finally picked up at the receding contact line. The particle layer at the liquid-air interface forms a mobile superhydrophobic layer itself in the case of hydrophobic particles. This principle was utilized to prepare uniquely shaped liquid marbles with polyhedral shapes like a cube. A droplet on a plate bed basically sits on a loose superhydrophobic layer and picks up the particles from this bed in a self-cleaning manner. Despite the random orien- tation of the hexagonal plates on such a bed, highly ordered two-dimensional layers were formed on the polyhedral liquid marbles. Such liquid marbles have the potential to ma- nipulate and transport small amounts of liquids or to be used as sensors since various gases can be exchanged through the gaps between the plates. They also present a novel model system to investigate crystallization on curved liquid surfaces. 146 REFERENCES 8 REFERENCES [1] P. G. de Gennes, Rev. Mod. Phys. 1985, 57, 827. [2] S. Herminghaus, Europhys. Lett. 2000, 52, 165. [3] D. Quéré Annu. Rev. Mater. Res. 2008, 38, 71. [4] D. Bonn, J. Eggers, J. Indekeu, J. Meunier, E. Rolley, Rev. Mod. Phys. 2009, 81, 739. [5] H.-J. Butt, K. Graf, M. Kappl, Physics and Chemistry of Interfaces, Wiley-VCH, Weinheim, Germany 2003. [6] G. Lagubeau, M. Le Merrer, C. Clanet, D. Quere, Nat. Phys. 2011, 7, 395. [7] D. W. Pilat, P. Papadopoulos, D. Schäffel, D. Vollmer, R. Berger, H. J. Butt, Langmuir 2012, 28, 16812. [8] D. ’t Mannetje, A. Banpurkar, H. Koppelman, M. H. G. Duits, D. van den Ende, F. Mugele, Langmuir 2013, 29, 9944. [9] R. Tadmor, P. Bahadur, A. Leh, H. E. N’guessan, R. Jaini, L. Dang, Phys. Rev. Lett. 2009, 103, 266101. [10] M. J. Kreder, J. Alvarenga, P. Kim, J. Aizenberg, Nat. Rev. Mater. 2016, 1, 15003. [11] I. Shestopalov, J. D. Tice, R. F. Ismagilov, Lab Chip 2004, 4, 316. [12] Y. Zheng, H. Bai, Z. Huang, X. Tian, F.-Q. Nie, Y. Zhao, J. Zhai, L. Jiang, Nature 2010, 463, 640. [13] K.-C. Park, S. S. Chhatre, S. Srinivasan, R. E. Cohen, G. H. McKinley, Langmuir 2013, 29, 13269. [14] P. Cheng, X. Quan, S. Gong, X. Liu, L. Yang, in Advances in Heat Transfer, Vol. 46 (Eds: Y. I. C. J. P. A. Ephraim M. Sparrow, G. John M), Elsevier 2014, p. 187. [15] K. Rykaczewski, A. T. Paxson, M. Staymates, M. L. Walker, X. Sun, S. Anand, S. Srinivasan, G. H. McKinley, J. Chinn, J. H. J. Scott, K. K. Varanasi, Sci. Rep. 2014, 4. [16] A. T. Paxson, J. L. Yagüe, K. K. Gleason, K. K. Varanasi, Adv. Mater. 2014, 26, 418. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 147 [17] C. Yang, U. Kaipa, Q. Z. Mather, X. Wang, V. Nesterov, A. F. Venero, M. A. Omary, J. Am. Chem. Soc. 2011, 133, 18094. [18] J. G. Nguyen, S. M. Cohen, J. Am. Chem. Soc. 2011, 132, 4560. [19] K. P. Rao, M. Higuchi, K. Sumida, S. Furukawa, J. Duan, S. Kitagawa, Angew. Chem. Int. Ed 2014, 126, 8364. [20] K. Jayaramulu, K. K. R. Datta, C. Rösler, M. Petr, M. Otyepka, R. Zboril, R. A. Fischer, Angew. Chem. Int. Ed. 2016, 55, 1178. [21] S. Roy, V. M. Suresh, T. K. Maji, Chem. Sci. 2016, 7, 2251. [22] Y. Zheng, X. Gao, L. Jiang, Soft Matter 2007, 3, 178. [23] Z. Guo, W. Liu, B.-L. Su, J. Colloid Interface Sci. 2011, 353, 335. [24] T. Young, Phil. Trans. R. Soc. Lond. 1805, 95, 65. [25] K.-Y. Law, J. Phys. Chem. Lett. 2014, 5, 686. [26] R. N. Wenzel, Ind. Eng. Chem. 1936, 28, 988. [27] A. B. D. Cassie, S. Baxter, Trans. Faraday Soc. 1944, 40, 546. [28] A. Lafuma, D. Quéré, Nat. Mater. 2003, 2, 457. [29] M. Callies, D. Quéré, Soft Matter 2005, 1, 55. [30] L. Feng, S. H. Li, Y. S. Li, H. J. Li, L. J. Zhang, J. Zhai, Y. L. Song, B. Q. Liu, L. Jiang, D. B. Zhu, Adv. Mater. 2002, 14, 1857. [31] B. Bhushan, Y. C. Jung, Prog. Mater. Sci. 2011, 56, 1. [32] Y. Y. Yan, N. Gao, W. Barthlott, Adv. Colloid Interface Sci. 2011, 169, 80. [33] R. H. Dettre, J. Johnson, Rulon E., in Contact Angle, Wettability, and Adhesion, DOI: 10.1021/ba-1964-0043.ch008 (Ed: F. M. Fowkes), American Chemical Society, Washington, D.C. 1964, Ch. 8, p. 136. [34] W. Barthlott, C. Neinhuis, Planta 1997, 202, 1. [35] X. Gao, L. Jiang, Nature 2004, 432, 36. [36] H. J. Ensikat, P. Ditsche-Kuru, C. Neinhuis, W. Barthlott, Beilstein J. Nanotechnol. 2011, 2, 152. 148 REFERENCES [37] L. Mammen, K. Bley, P. Papadopoulos, F. Schellenberger, N. Encinas, H.-J. Butt, C. K. Weiss, D. Vollmer, Soft Matter 2014, 11, 506. [38] M. D'Acunzi, L. Mammen, M. Singh, X. Deng, M. Roth, G. K. Auernhammer, H.-J. Butt, D. Vollmer, Faraday Discuss. 2010, 146, 35. [39] A. Marmur, Langmuir 2004, 20, 3517. [40] H. Teisala, M. Tuominen, M. Aromaa, J. M. Mäkelä, M. Stepien, J. J. Saarinen, M. Toivakka, J. Kuusipalo, Surf. Coat. Technol. 2010, 205, 436. [41] X. Deng, L. Mammen, H.-J. Butt, D. Vollmer, Science 2012, 335, 67. [42] E. Martines, K. Seunarine, H. Morgan, N. Gadegaard, C. D. W. Wilkinson, M. O. Riehle, Nano Lett. 2005, 5, 2097. [43] D. Öner, T. J. McCarthy, Langmuir 2000, 16, 7777. [44] B. Qian, Z. Shen, Langmuir 2005, 21, 9007. [45] S. Minko, M. Müller, M. Motornov, M. Nitschke, K. Grundke, M. Stamm, J. Am. Chem. Soc. 2003, 125, 3896. [46] X. Wang, B. Ding, J. Yu, M. Wang, Nano Today 2011, 6, 510. [47] K. S. Birdi, CRC Press, Boca Raton 2015. [48] K. S. Birdi, Handbook of Surface and Colloid Chemistry, CRC Press, Boca Raton 2015. [49] P. L. du Noüy, J. Gen. Physiol. 1925, 7, 625. [50] L. Wilhelmy, Pogg. Ann. 1863, 119, 177. [51] W. A. Zisman, in Contact Angle, Wettability, and Adhesion (Ed: F. M. Fowkes), American Chemical Society, Washington, D.C. 1964, Ch. 1, p. 1. [52] F. M. Fowkes, Ind. Eng. Chem. 1964, 56, 40. [53] D. K. Owens, R. C. Wendt, J. Appl. Polym. Sci. 1969, 13, 1741. [54] W. Rabel, Farbe und Lacke 1971, 77, 997. [55] D. H. Kaelble, J. Adhesion 1970, 2, 66. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 149 [56] N. L. Jarvis, W. A. Zisman, Surface Chemistry of Fluorochemicals, U.S. Naval Research Laboratory, Washington, D.C. 1965. [57] J. J. Gilman, J. Appl. Phys. 1960, 31, 2208. [58] T. Nishino, M. Meguro, K. Nakamae, M. Matsushita, Y. Ueda, Langmuir 1999, 15, 4321. [59] T. Huhtamäki, X. Tian, J. T. Korhonen, R. H. A. Ras, Nat. Protoc. 2018, 13, 1521. [60] X. Zhang, F. Shi, J. Niu, Y. Jiang, Z. Wang, J. Mater. Chem. 2008, 18, 621. [61] F. Schellenberger, N. Encinas, D. Vollmer, H.-J. Butt, Phys. Rev. Lett. 2016, 116, 096101. [62] P. Papadopoulos, L. Mammen, X. Deng, D. Vollmer, H.-J. Butt, Proc. Natl. Acad. Sci. USA 2013, 110, 3254. [63] L. Gao, T. J. McCarthy, Langmuir 2006, 22, 6234. [64] H.-J. Butt, R. Berger, W. Steffen, D. Vollmer, S. A. L. Weber, Langmuir 2018, 34, 11292. [65] R. H. A. Ras, X. Tian, I. S. Bayer, in Handbook of Nanocellulose and Cellulose Nanocomposites, Vol. 1 (Eds: H. Kargarzadeh, I. Ahmad, S. Thomas, A. Dufresne), Wiley-VCH, Weinheim, Germany 2017, Ch. 22. [66] C. J. Long, J. F. Schumacher, A. B. Brennan, Langmuir 2009, 25, 12982. [67] L. Gao, T. J. McCarthy, J. Am. Chem. Soc. 2006, 128, 9052. [68] L. Wang, T. J. McCarthy, Angew. Chem. Int. Ed. 2016, 55, 244. [69] M. Miwa, A. Nakajima, A. Fujishima, K. Hashimoto, T. Watanabe, Langmuir 2000, 16, 5754. [70] F. Schellenberger, J. Xie, N. Encinas, A. Hardy, M. Klapper, P. Papadopoulos, H.- J. Butt, D. Vollmer, Soft Matter 2015, 11, 7617. [71] T.-S. Wong, S. H. Kang, S. K. Y. Tang, E. J. Smythe, B. D. Hatton, A. Grinthal, J. Aizenberg, Nature 2011, 477, 443. [72] Y. I. Frenkel, J. Exp. Theor. Phys. 1948, 18, 658. [73] K. Kawasaki, J. Colloid Sci. 1960, 15, 402. 150 REFERENCES [74] R. A. Brown, F. M. Orr Jr., L. E. Scriven, J. Colloid Interface Sci. 1980, 73, 76. [75] C. W. Extrand, A. N. Gent, J. Colloid Interface Sci. 1990, 138, 431. [76] C. W. Extrand, Y. Kumagai, J. Colloid Interface Sci. 1995, 170, 515. [77] N. Gao, F. Geyer, D. W. Pilat, S. Wooh, D. Vollmer, H.-J. Butt, R. Berger, Nat. Phys. 2018, 14, 191. [78] L. Feng, Z. Zhang, Z. Mai, Y. Ma, B. Liu, L. Jiang, D. Zhu, Angew. Chem. Int. Ed. 2004, 43, 2012. [79] J. Zhang, S. Seeger, Adv. Funct. Mater. 2011, 21, 4699. [80] Z. Chu, Y. Feng, S. Seeger, Angew. Chem. Int. Ed. 2015, 54, 2328. [81] K. Tsujii, T. Yamamoto, T. Onda, S. Shibuichi, Angew. Chem., Int. Ed. 1997, 36, 1011. [82] A. Tuteja, W. Choi, M. Ma, J. M. Mabry, S. A. Mazzella, G. C. Rutledge, G. H. McKinley, R. E. Cohen, Science 2007, 318, 1618. [83] A. K. Kota, G. Kwon, A. Tuteja, NPG Asia Mater. 2014, 6, e109. [84] M. Paven, P. Papadopoulos, S. Schöttler, X. Deng, V. Mailänder, D. Vollmer, H.-J. Butt, Nat. Commun. 2013, 4, 2512. [85] S. Movafaghi, V. Leszczak, W. Wang, J. A. Sorkin, L. P. Dasi, K. C. Popat, A. K. Kota, Adv. Healthcare Mater. 2017, 6, 1600717. [86] A. Tuteja, W. Choi, J. M. Mabry, G. H. McKinley, R. E. Cohen, Proc. Natl. Acad. Sci. USA 2008, 105, 18200. [87] H.-J. Butt, C. Semprebon, P. Papadopoulos, D. Vollmer, M. Brinkmann, M. Ciccotti, Soft Matter 2013, 9, 418. [88] Z. Chu, S. Seeger, Chem. Soc. Rev. 2014, 43, 2784. [89] S. Peng, X. Yang, D. Tian, W. Deng, ACS Appl. Mater. Interfaces 2014, 6, 15188. [90] D. Wang, X. Wang, X. Liu, F. Zhou, J. Phys. Chem. C 2010, 114, 9938. [91] W. Wu, X. Wang, D. Wang, M. Chen, F. Zhou, W. Liu, Q. Xue, Chem. Commun. 2009, DOI: 10.1039/B818633B1043. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 151 [92] L. Li, V. Breedveld, D. W. Hess, ACS Appl. Mater. Interfaces 2013, 5, 5381. [93] A. Ahuja, J. A. Taylor, V. Lifton, A. A. Sidorenko, T. R. Salamon, E. J. Lobaton, P. Kolodner, T. N. Krupenkin, Langmuir 2008, 24, 9. [94] H. Teisala, F. Geyer, J. Haapanen, P. Juuti, J. M. Mäkelä, D. Vollmer, H. J. Butt, Adv. Mater. 2018, 30, 1706529. [95] K. Golovin, D. H. Lee, J. M. Mabry, A. Tuteja, Angew. Chem. Int. Ed. 2013, 52, 13007. [96] S. Pan, R. Guo, M. Björnmalm, J. J. Richardson, L. Li, C. Peng, N. Bertleff- Zieschang, W. Xu, J. Jiang, F. Caruso, Nat. Mater. 2018, 17, 1040. [97] H. Zhou, H. Wang, H. Niu, A. Gestos, T. Lin, Adv. Funct. Mater. 2013, 23, 1664. [98] W. Choi, A. Tuteja, S. Chhatre, J. M. Mabry, R. E. Cohen, G. H. McKinley, Adv. Mater. 2009, 21, 2190. [99] C.-T. Hsieh, F.-L. Wu, W.-Y. Chen, J. Phys. Chem. C, 2009, 113, 13683. [100] S. Pan, A. K. Kota, J. M. Mabry, A. Tuteja, J. Am. Chem. Soc. 2013, 135, 578. [101] T. Darmanin, F. Guittard, S. Amigoni, E. T. de Givenchy, X. Noblin, R. Kofman, F. Celestini, Soft Matter 2010, 7, 1053. [102] N. J. Shirtcliffe, G. McHale, M. I. Newton, Y. Zhang, ACS Appl. Mater. Interfaces 2009, 1, 1316. [103] Z. Yu, X. Liu, G. Kuang, Chinese J. Chem. Eng. 2015, 23, 763. [104] C. Xiang, L. Sun, Y. Wang, G. Wang, X. Zhao, S. Zhang, J. Phys. Chem. C 2017, 121, 15448. [105] C. Q. Lai, J. C. W. Shen, W. C. W. Cheng, C. H. Yap, RSC Adv. 2016, 6, 62451. [106] Y. W. Kim, J. M. Lee, I. Lee, S. H. Lee, J. S. Ko, J. Precis. Eng. Manuf. 2013, 14, 299. [107] B. H. Kwon, H. H. Kim, H. J. Jeon, M. C. Kim, I. Lee, S. Chun, J. S. Go, Exp. Fluids 2014, 55, 1722. [108] J. M. Nokes, R. Liedert, M. Y. Kim, A. Siddiqui, M. Chu, E. K. Lee, M. Khine, Adv. Healthcare Mater. 2016, 5, 593. [109] D. Kim, W. Hwang, J. Micromech. Microeng. 2010, 20, 027002. 152 REFERENCES [110] S. Hoshian, E. Kankuri, R. H. A. Ras, S. Franssila, V. Jokinen, Sci. Rep. 2017, 7, 16019. [111] N. R. Geraldi, L. E. Dodd, B. B. Xu, G. G. Wells, D. Wood, M. I. Newton, G. McHale, Surf. Topogr.: Metrol. Prop. 2017, 5, 034001. [112] G. R. J. Artus, S. Jung, J. Zimmermann, H.-P. Gautschi, K. Marquardt, S. Seeger, Adv. Mater. 2006, 18, 2758. [113] J. Zimmermann, F. A. Reifler, G. Fortunato, L. C. Gerhardt, S. Seeger, Adv. Funct. Mater. 2008, 18, 3662. [114] J. Zhang, L. Li, B. Li, S. Seeger, J. Zhang, L. Li, B. Li, S. Seeger, RSC Adv. 2014, 4, 33424. [115] J. Zhang, S. Seeger, Angew. Chem. Int. Ed. 2011, 50, 6652. [116] F. Geyer, C. Schönecker, H.-J. Butt, D. Vollmer, Adv. Mater. 2017, 29, 1603524. [117] J. Zhang, L. Li, B. Li, S. Seeger, RSC Adv. 2014, 4, 33424. [118] F. Geyer, M. D'Acunzi, C. Y. Yang, M. Müller, P. Baumli, A. Kaltbeitzel, V. Mailänder, N. Encinas, D. Vollmer, H. J. Butt, Adv. Mater. 2019, 31, 1801324. [119] D.-a. E. Rollings, J. G. C. Veinot, Langmuir 2008, 24, 13653. [120] D.-a. E. Rollings, S. Tsoi, J. C. Sit, J. G. C. Veinot, Langmuir 2007, 23, 5275. [121] M. Jin, J. Wang, Y. Hao, M. Liao, Y. Zhao, Polym. Chem. 2011, 2, 1658. [122] J. X. H. Wong, H. Asanuma, H.-Z. Yu, Thin Solid Films 2012, 522, 159. [123] J. Zhang, A. Wang, S. Seeger, Polym. Chem. 2014, 5, 1132. [124] L. Gao, T. J. McCarthy, Langmuir 2008, 24, 362. [125] J. Zimmermann, G. R. J. Artus, S. Seeger, Appl. Surf. Sci. 2007, 253, 5972. [126] J. Zimmermann, F. A. Reifler, U. Schrade, G. R. J. Artus, S. Seeger, Colloid Surface A 2007, 302, 234. [127] G. R. J. Artus, S. Seeger, Adv. Colloid Interface Sci. 2014, 209, 144. [128] G. R. J. Artus, S. Olveira, D. Patra, S. Seeger, Macromol. Rapid Commun. 2017, 38, 1600558. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 153 [129] S. Olveira, A. Stojanovic, S. Seeger, J. Nanopart. Res. 2018, 20, 307. [130] G. R. J. Artus, S. Seeger, Advances in Colloid and Interface Science 2014, 209, 144. [131] J. T. Korhonen, T. Huhtamäki, T. Verho, R. H. A. Ras, Surf. Innov. 2014, 2, 116. [132] J. Zimmermann, M. Rabe, G. R. J. Artus, S. Seeger, Soft Matter 2008, 4, 450. [133] G. R. J. Artus, J. Zimmermann, F. A. Reifler, S. A. Brewer, S. Seeger, Appl. Surf. Sci. 2012, 258, 3835. [134] X. Deng, M. Paven, P. Papadopoulos, M. Ye, S. Wu, T. Schuster, M. Klapper, D. Vollmer, H.-J. Butt, Angew. Chem. Int. Ed. 2013, 52, 11286. [135] A. M. S. Costa, M. Alatorre-Meda, N. M. Oliveira, J. F. Mano, Langmuir 2014, 30, 4535. [136] S. Wooh, H. Huesmann, M. N. Tahir, M. Paven, K. Wichmann, D. Vollmer, W. Tremel, P. Papadopoulos, H.-J. Butt, Adv. Mater. 2015, 27, 7338. [137] T. Sekido, S. Wooh, R. Fuchs, M. Kappl, Y. Nakamura, H.-J. Butt, S. Fujii, Langmuir 2017, 33, 1995. [138] J. T. Pham, M. Paven, S. Wooh, T. Kajiya, H.-J. Butt, D. Vollmer, Nat. Commun. 2017, 8, 905. [139] B. Deng, R. Cai, Y. Yu, H. Q. Jiang, C. L. Wang, J. A. Li, L. F. Li, M. Yu, J. Y. Li, L. D. Xie, Q. Huang, C. H. Fan, Adv. Mater. 2010, 22, 5473. [140] Y. Shang, Y. Si, A. Raza, L. Yang, X. Mao, B. Ding, J. Yu, Nanoscale 2012, 4, 7847. [141] Y. Si, Q. Fu, X. Wang, J. Zhu, J. Yu, G. Sun, B. Ding, ACS Nano 2015, 9, 3791. [142] X. Zhou, Z. Zhang, X. Xu, F. Guo, X. Zhu, X. Men, B. Ge, ACS Appl. Mater. Interfaces 2013, 5, 7208. [143] L. Cao, A. K. Jones, V. K. Sikka, J. Wu, D. Gao, Langmuir 2009, 25, 12444. [144] S. A. Kulinich, M. Farzaneh, Appl. Surf. Sci. 2009, 255, 8153. [145] R. Fürstner, W. Barthlott, C. Neinhuis, P. Walzel, Langmuir 2005, 21, 956. [146] B. Bhushan, Y. C. Jung, K. Koch, Langmuir 2009, 25, 3240. [147] R. J. Daniello, N. E. Waterhouse, J. P. Rothstein, Phys. Fluids 2009, 21, 085103. 154 REFERENCES [148] B. J. Privett, J. Youn, S. A. Hong, J. Lee, J. Han, J. H. Shin, M. H. Schoenfisch, Langmuir 2011, 27, 9597. [149] Y. Liao, R. Wang, A. G. Fane, J. Membrane Sci. 2013, 440, 77. [150] A. Razmjou, E. Arifin, G. Dong, J. Mansouri, V. Chen, J. Membr. Sci. 2012, 415, 850. [151] U. Pöschl, Angew. Chem. Int. Ed. 2005, 44, 7520. [152] X.-M. Li, D. Reinhoudt, M. Crego-Calama, Chem. Soc. Rev. 2007, 36, 1350. [153] B. Bhushan, Y. C. Jung, K. Koch, Langmuir 2009, 25, 3240. [154] Y.-B. Park, H. Im, M. Im, Y.-K. Choi, J. Mater. Chem. 2011, 21, 633. [155] K. M. Wisdom, J. A. Watson, X. Qu, F. Liu, G. S. Watson, C.-H. Chen, Proc. Natl. Acad. Sci. USA 2013, 110, 7992. [156] M. Jönsson-Niedziólka, F. Lapierre, Y. Coffinier, S. J. Parry, F. Zoueshtiagh, T. Foat, V. Thomy, R. Boukherroub, Lab Chip 2011, 11, 490. [157] S. Wooh, H. Huesmann, M. N. Tahir, M. Paven, K. Wichmann, D. Vollmer, W. Tremel, P. Papadopoulos, H. Butt, Adv. Mater. 2015, 27, 7338. [158] Q. Zhu, Q. Pan, F. Liu, J. Phys. Chem. C 2011, 115, 17464. [159] R. B. Pernites, C. M. Santos, M. Maldonado, R. R. Ponnapati, D. F. Rodrigues, R. C. Advincula, Chem. Mater. 2012, 24, 870. [160] C.-Y. Loo, P. M. Young, W.-H. Lee, R. Cavaliere, C. B. Whitchurch, R. Rohanizadeh, Acta Biomater. 2012, 8, 1881. [161] X. X. Zhang, L. Wang, E. Levanen, RSV Adv. 2013, 3, 12003. [162] X.-Q. Dou, D. Zhang, C. Feng, L. Jiang, ACS Nano 2015, 9, 10664. [163] G. B. Hwang, K. Page, A. Patir, S. P. Nair, E. Allan, I. P. Parkin, ACS Nano 2018, 12, 6050. [164] M. Meier, V. Dubois, S. Seeger, Appl. Surf. Sci. 2018, 459, 505. [165] J. W. Costerton, P. S. Stewart, E. P. Greenberg, Science 1999, 284, 1318. [166] L. D. Renner, D. B. Weibel, MRS Bull. 2011, 36, 347. [167] P. S. Stewart, J. W. Costerton, Lancet 2001, 358, 135. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 155 [168] V. K. Truong, R. Lapovok, Y. S. Estrin, S. Rundell, J. Y. Wang, C. J. Fluke, R. J. Crawford, E. R. Ivanova, Biomaterials 2010, 31, 3674. [169] S. J. Cho, H. Nam, H. Ryu, G. Lim, Adv. Funct. Mater. 2013, 23, 5577. [170] T. C. Merkel, H. Lin, X. Wei, R. Baker, J. Membr. Sci. 2010, 359, 126. [171] P. H. M. Feron, A. E. Jansen, R. Klaassen, Energy Convers. Manage. 1992, 33, 421. [172] A. Gabelman, S.-T. Hwang, J. Membr. Sci. 1999, 159, 61. [173] A. Mansourizadeh, A. F. Ismail, J. Hazard. Mater. 2009, 171, 38. [174] P. Luis, T. Van Gerven, B. Van der Bruggen, Prog. Energy Combust. Sci. 2012, 38, 419. [175] M. T. Ravanchi, T. Kaghazchi, A. Kargari, Desalination 2009, 235, 199. [176] R. Wang, H. Y. Zhang, P. H. M. Feron, D. T. Liang, Sep. Purif. Technol. 2005, 46, 33. [177] G. T. Rochelle, Science 2009, 325, 1652. [178] R. S. Haszeldine, Science 2009, 325, 1647. [179] B. P. Binks, R. Murakami, Nat. Mater. 2006, 5, 865. [180] S. U. Pickering, J. Chem. Soc., Trans. 1907, 91, 2001. [181] B. P. Binks, T. S. Horozov, Colloidal Particles at Liquid Interfaces, Cambridge University Press, Cambridge 2006. [182] S. Fujii, Y. Nakamura, Langmuir 2017, 33, 7365. [183] S. Fujii, Y. Yokoyama, S. Nakayama, M. Ito, S. Yusa, Y. Nakamura, Langmuir 2017, 34, 933. [184] G. McHale, M. I. Newton, Soft Matter 2011, 7, 5473. [185] P. Aussillous, D. Quéré, Nature 2001, 411, 924. [186] P. Aussillous, D. Quéré, Proc. R. Soc. A 2006, 462, 973. [187] G. McHale, M. I. Newton, Soft Matter 2015, 11, 2530. [188] E. Bormashenko, Langmuir 2017, 33, 663. [189] N. Pike, D. Richard, W. Foster, L. Mahadevan, Proc. Biol. Sci. 2002, 269, 1211. [190] D. Dupin, S. P. Armes, S. Fujii, J. Am. Chem. Soc. 2009, 131, 5386. 156 REFERENCES [191] P. M. Ireland, C. A. Thomas, B. T. Lobel, G. B. Webber, S. Fujii, E. J. Wanless, Front. Chem. 2018, 6, 280. [192] D. Matsukuma, H. Watanabe, M. Minn, A. Fujimoto, T. Shinohara, H. Jinnai, A. Takahara, RSC Adv. 2013, 3, 7862. [193] S. Fujii, S. Sawada, S. Nakayama, M. Kappl, K. Ueno, K. Shitajima, H.-J. Butt, Y. Nakamura, Mater. Horiz. 2016, 3, 42. [194] Y. Sheng, G. Sun, J. Wu, G. Ma, T. Ngai, Angew. Chem. Int. Ed. 2015, 54, 7012. [195] A. T. Tyowua, J. M. Mooney, B. P. Binks, Colloids Surf. A 2019, 560, 288. [196] A. D. Dinsmore, M. F. Hsu, M. G. Nikolaides, M. Marquez, A. R. Bausch, D. A. Weitz, Science 2002, 298, 1006. [197] S. Fujii, S.-i. Yusa, Y. Nakamura, Adv. Funct. Mater. 2016, 26, 7206. [198] T. Arbatan, L. Li, J. Tian, W. Shen, Adv. Healthcare Mater. 2012, 1, 80. [199] Y. Xue, H. Wang, Y. Zhao, L. Dai, L. Feng, X. Wang, T. Lin, Adv. Mater. 2010, 22, 4814. [200] E. Bormashenko, R. Pogreb, Y. Bormashenko, A. Musin, T. Stein, Langmuir 2008, 24, 12119. [201] N. M. Oliveira, R. L. Reis, J. F. Mano, Adv. Healthcare Mater. 2017, 6, 1700192. [202] N. M. Oliveira, C. R. Correia, R. L. Reis, J. F. Mano, Adv. Healthcare Mater. 2015, 4, 264. [203] H. Zeng, Y. Zhao, Appl. Phys. Lett. 2010, 96, 114104. [204] J. F. Archard, J. Appl. Phys. 1953, 24, 981. [205] B. T. Persson, in Encyclopedia of Lubricants and Lubrication, DOI: 10.1007/978-3- 642-22647-2_80 (Ed: T. Mang), Springer Berlin Heidelberg 2014, Ch. 80, p. 791. [206] H.-J. Butt, M. Kappl, Friction, Wiley-VCH, Weinheim, Germany 2010. [207] P. Calvert, Chem. Mater. 2001, 13, 3299. [208] F. P. Bowden, D. Tabor, The Friction and Lubrication of Solids, Clarendon Press, 2001. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 157 [209] B. Bhushan, J. N. Israelachvili, U. Landman, Nature 1995, 374, 607. [210] C. Semprebon, M. Brinkmann, Soft Matter 2014, 10, 3325. [211] C. W. Extrand, A. N. Gent, J. Colloid Interface Sci. 1990, 138, 431. [212] R. A. Brown, F. M. Orr Jr, L. E. Scriven, J. Colloid Interface Sci. 1980, 73, 76. [213] C. W. Extrand, Y. Kumagai, J. Colloid Interface Sci. 1995, 170, 515. [214] Y. I. Frenkel, J. Exp. Theor. Phys. 1948, 18, 658. [215] K. Kawasaki, J. Colloid Sci. 1960, 15, 402. [216] A. ElSherbini, A. Jacobi, J. Colloid Interface Sci. 2006, 299, 841. [217] V. Dussan, J. Fluid Mech. 1987, 174, 381. [218] E. Wolfram, R. Faust, in Wetting, Spreading, and Adhesion (Ed: J. F. Padday), Academic Press, New York 1978, p. 213. [219] C. Antonini, F. J. Carmona, E. Pierce, M. Marengo, A. Amirfazli, Langmuir 2009, 25, 6143. [220] V. Berejnov, R. E. Thorne, Phys. Rev. E 2007, 75, 066308. [221] J. V. I. Timonen, M. Latikka, O. Ikkala, R. H. A. Ras, Nat. Commun. 2013, 4. [222] P. Olin, S. B. Lindström, T. Pettersson, L. Wågberg, Langmuir 2013, 29, 9079. [223] E. Pierce, F. J. Carmona, A. Amirfazli, Colloid Surf. A-Physicochem. Eng. Asp. 2008, 323, 73. [224] M. Sakai, J.-H. Song, N. Yoshida, S. Suzuki, Y. Kameshima, A. Nakajima, Langmuir 2006, 22, 4906. [225] P. R. Griffiths, Master of Science in Mechanical Engineering University of South Florida, 2013. [226] J. H. Snoeijer, B. Andreotti, Annu. Rev. Fluid Mech. 2013, 45, 269. [227] H. Perrin, R. Lhermerout, K. Davitt, E. Rolley, B. Andreotti, Phys. Rev. Lett. 2016, 116, 184502. [228] K. Huang, I. Szlufarska, Phys. Rev. E 2014, 89, 032119. [229] D. Daniel, J. V. I. Timonen, R. Li, S. J. Velling, J. Aizenberg, Nat. Phys. 2017, advance online publication. 158 REFERENCES [230] H.-J. Butt, N. Gao, P. Papadopoulos, W. Steffen, M. Kappl, R. Berger, Langmuir 2017, 33, 107. [231] J. N. Israelachvili, Intermolecular and Surface Forces, Elsevier Science, 2011. [232] H.-J. Butt, M. Kappl, Surface and Interfacial Forces, Wiley, 2010. [233] C. Shi, X. Cui, X. Zhang, P. Tchoukov, Q. Liu, N. Encinas, M. Paven, F. Geyer, D. Vollmer, Z. Xu, H.-J. Butt, H. Zeng, Langmuir 2015, 31, 7317. [234] J. W. Krumpfer, T. J. McCarthy, Farad. Discuss. 2010, 146, 103. [235] S. Wooh, J. H. Koh, S. Lee, H. Yoon, K. Char, Adv. Funct. Mater. 2014, 24, 5550. [236] H. Suda, S. Yamada, Langmuir 2003, 19, 529. [237] K. L. Johnson, K. L. Johnson, Contact Mechanics, Cambridge University Press, 1987. [238] T. Darmanin, F. Guittard, J. Mater. Chem. A 2014, 2, 16319. [239] L. Mishchenko, B. Hatton, V. Bahadur, J. A. Taylor, T. Krupenkin, J. Aizenberg, ACS Nano 2010, 4, 7699. [240] R. P. Garrod, L. G. Harris, W. C. E. Schofield, J. McGettrick, L. J. Ward, a. D. O. H. Teare, J. P. S. Badyal*, Langmuir 2006, 23, 689. [241] K.-C. Park, S. S. Chhatre, S. Srinivasan, R. E. Cohen, G. H. McKinley, Langmuir 2013, 29, 13269. [242] P. Saxena, L. M. Hildemann, J. Atmos. Chem. 1996, 24, 57. [243] A. P. Sullivan, R. J. Weber, A. L. Clements, J. R. Turner, M. S. Bae, J. J. Schauer, Geophys. Res. Lett. 2004, 31, L13105. [244] R. D. Deegan, O. Bakajin, T. F. Dupont, G. Huber, S. R. Nagel, T. A. Witten, Nature 1997, 389, 827. [245] O. Pitois, X. Chateau, Langmuir 2002, 18, 9751. [246] Y. Tang, S. Cheng, Phys. Rev. E 2018, 98, 032802. [247] K. L. Johnson, K. Kendall, A. D. Roberts, Proc. R. Soc. Lond. A 1971, 324, 301. [248] J. B. Boreyko, C.-H. Chen, Phys. Rev. Lett. 2009, 103, 184501. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 159 [249] M. D'Acunzi, PhD thesis Johannes Gutenberg-University, Mainz, Mainz, 2010. [250] N. A. M. Verhaegh, A. van Blaaderen, 1994, 10, 1427. [251] H. J. H. Fenton, J. Chem. Soc., Trans. 1894, 65, 899. [252] J. J. Pignatello, E. Oliveros, A. MacKay, Crit. Rev. Environ. Sci. Technol. 2006, 36, 1. [253] L. Zhang, M. D'Acunzi, M. Kappl, G. K. Auernhammer, D. Vollmer, C. M. van Kats, A. van Blaaderen, Langmuir 2009, 25, 2711. [254] W. Schaertl, H. Sillescu, J. Colloid Interface Sci. 1993, 155, 313. [255] C. Huh, S. G. Mason, Can. J. Chem. 1976, 54, 969. [256] E. Bayramli, C. Huh, S. G. Mason, Can. J. Chem. 1978, 56, 818. [257] O. D. Velev, K. Furusawa, K. Nagayama, Langmuir 1996, 12, 2374. [258] S. E. Anachkov, I. Lesov, M. Zanini, P. A. Kralchevsky, N. D. Denkov, L. Isa, Soft Matter 2016, 12, 7632. [259] H. H. Kausch, D. G. Fesko, N. W. Tschoegl, J. Colloid Interface Sci. 1971, 37, 603. [260] P. M. Cox, R. A. Betts, C. D. Jones, S. A. Spall, I. J. Totterdell, Nature 2000, 408, 184. [261] C. L. Quéré, R. Moriarty, R. M. Andrew, G. P. Peters, P. Ciais, P. Friedlingstein, S. D. Jones, S. Sitch, P. Tans, A. Arneth, T. A. Boden, L. Bopp, Y. Bozec, J. G. Canadell, L. P. Chini, F. Chevallier, C. E. Cosca, I. Harris, M. Hoppema, R. A. Houghton, J. I. House, A. K. Jain, T. Johannessen, E. Kato, R. F. Keeling, V. Kitidis, K. Klein Goldewijk, C. Koven, C. S. Landa, P. Landschützer, A. Lenton, I. D. Lima, G. Marland, J. T. Mathis, N. Metzl, Y. Nojiri, A. Olsen, T. Ono, S. Peng, W. Peters, B. Pfeil, B. Poulter, M. R. Raupach, P. Regnier, C. Rödenbeck, S. Saito, J. E. Salisbury, U. Schuster, J. Schwinger, R. Séférian, J. Segschneider, T. Steinhoff, B. D. Stocker, A. J. Sutton, T. Takahashi, B. Tilbrook, G. R. v. d. Werf, N. Viovy, Y.-P. Wang, R. Wanninkhof, A. Wiltshire, N. Zeng, Earth Syst. Sci. Data 2015, 7, 47. [262] D. M. D'Alessandro, B. Smit, J. R. Long, Angew. Chem. Int. Ed. 2010, 49, 6058. [263] S. J. Datta, C. Khumnoon, Z. H. Lee, W. K. Moon, S. Docao, T. H. Nguyen, I. C. Hwang, D. Moon, P. Oleynikov, O. Terasaki, K. B. Yoon, Science 2015, 350, 302. 160 REFERENCES [264] H. W. Kim, H. W. Yoon, S. M. Yoon, B. M. Yoo, B. K. Ahn, Y. H. Cho, H. J. Shin, H. Yang, U. Paik, S. Kwon, J. Y. Choi, H. B. Park, Science 2013, 342, 91. [265] S. A. Freeman, R. Dugas, D. H. Van Wagener, T. Nguyen, G. T. Rochelle, Int. J. Greenh. Gas Con. 2010, 4, 119. [266] P. Luis, B. Van der Bruggen, T. Van Gerven, J. Chem. Technol. Biotechnol. 2011, 86, 769. [267] S.-W. Rho, K.-P. Yoo, J. S. Lee, S. C. Nam, J. E. Son, B.-M. Min, J. Chem. Eng. Data 1997, 42, 1161. [268] A. Criscuoli, E. Drioli, Ind. Eng. Chem. Res. 2007, 46, 2268. [269] A. Stankiewicz, Chem. Eng. Process. 2003, 42, 137. [270] T. Van Gerven, A. Stankiewicz, Ind. Eng. Chem. Res. 2009, 48, 2465. [271] E. Drioli, A. I. Stankiewicz, F. Macedonio, J. Membr. Sci. 2011, 380, 1. [272] Y. Lv, X. Yu, J. Jia, S.-T. Tu, J. Yan, E. Dahlquist, Appl. Energy 2012, 90, 167. [273] Y.-F. Lin, Q. Ye, S.-H. Hsu, T.-W. Chung, Chem. Eng. J. 2016, 284, 888. [274] X. Yu, L. An, J. Yang, S.-T. Tu, J. Yan, J. Membr. Sci. 2015, 496, 1. [275] T. Liu, C.-J. Kim, Science 2014, 346, 1096. [276] H. A. Al-Ghawas, D. P. Hagewiesche, G. Ruiz-Ibanez, O. C. Sandall, J. Chem. Eng. Data 1989, 34, 385. [277] G. Richner, Energy Procedia 2013, 37, 423. [278] C. W. Extrand, Langmuir 2002, 18, 7991. [279] L. Gao, T. J. McCarthy, Langmuir 2007, 23, 3762. [280] W. Choi, A. Tuteja, J. M. Mabry, R. E. Cohen, G. H. McKinley, J. Colloid Interface Sci. 2009, 339, 208. [281] P. S. H. Forsberg, C. Priest, M. Brinkmann, R. Sedev, J. Ralston, Langmuir 2010, 26, 860. [282] B. M. Mognetti, J. M. Yeomans, Langmuir 2010, 26, 18162 [283] C. W. Extrand, Langmuir 2016, 32, 8608. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 161 [284] P. Papadopoulos, X. Deng, L. Mammen, D.-M. Drotlef, G. Battagliarin, C. Li, K. Muellen, K. Landfester, A. del Campo, H.-J. Butt, D. Vollmer, Langmuir 2012, 28, 8392. [285] E. P. Barrett, L. G. Joyner, P. P. Halenda, J. Am. Chem. Soc. 1951, 73, 373. [286] S. Brunauer, P. H. Emmett, E. Teller, J. Am. Chem. Soc. 1938, 60, 309. [287] F. Y. Jou, A. E. Mather, F. D. Otto, Ind. Eng. Chem. Process. Des. Dev. 1982, 21, 539. [288] D. Menne, F. Pitsch, J. E. Wong, A. Pich, M. Wessling, Angew. Chem. Int. Ed. 2014, 53, 5706. [289] H. A. Stone, A. D. Stroock, A. Ajdari, Annu. Rev. Fluid Mech. 2004, 36, 381. [290] E. H. Kass, L. J. Schneiderman, N. Engl. J. Med. 1957, 256, 556. [291] X. J. Feng, L. Jiang, 2006, 18, 3063. [292] D. C. Leslie, A. Waterhouse, J. B. Berthet, T. M. Valentin, A. L. Watters, A. Jain, P. Kim, B. D. Hatton, A. Nedder, K. Donovan, E. H. Super, C. Howell, C. P. Johnson, T. L. Vu, D. E. Bolgen, S. Rifai, A. R. Hansen, M. Aizenberg, M. Super, J. Aizenberg, D. E. Ingber, Nat. Biotechnol, 2014, 32, 1134. [293] T. Verho, C. Bower, P. Andrew, S. Franssila, O. Ikkala, R. H. A. Ras, Adv. Mater. 2011, 23, 673. [294] X. Yao, Y. Song, L. Jiang, Adv. Mater. 2011, 23, 719. [295] C. W. Extrand, Langmuir 2004, 20, 5013. [296] D. Quéré, Rep. Prog. Phys. 2005, 68, 2495. [297] S. Herminghaus, EPL 2007, 79, 59901. [298] A. T. Paxson, K. K. Varanasi, Nat. Commun. 2013, 4, 1492. [299] H. Teisala, M. Tuominen, M. Aromaa, M. Stepien, J. M. Mäkelä, J. J. Saarinen, M. Toivakka, J. Kuusipalo, Langmuir 2012, 28, 3138. [300] M. R. Sanchis, V. Blanes, M. Blanes, D. Garcia, R. Balart, Eur. Polym. J. 2006, 42, 1558. [301] X. Tian, T. Verho, R. H. A. Ras, Science 2016, 352, 142. 162 REFERENCES [302] A. B. Subramaniam, M. Abkarian, H. A. Stone, Nat. Mater. 2005, 4, 553. [303] K. Kido, P. M. Ireland, T. Sekido, E. J. Wanless, G. B. Webber, Y. Nakamura, S. Fujii, Langmuir 2018, 34, 4970. [304] S. Wong, V. Kitaev, G. A. Ozin, J. Am. Chem. Soc. 2003, 125, 15589. [305] H. W. Deckman, J. H. Dunsmuir, S. Garoff, J. A. McHenry, D. G. Peiffer, J. Vac. Sci. Technol. B 1998, 6, 333. [306] N. Vogel, J. Ally, K. Bley, M. Kappl, K. Landfester, C. K. Weiss, Nanoscale 2014, 6, 6879. [307] Y. Nonomura, S. Komura, K. Tsujii, Langmuir 2004, 20, 11821. [308] R. G. Picknett, R. Bexon, J. Colloid Interface Sci. 1977, 61, 336. [309] B. Laborie, F. Lachaussée, E. Lorenceau, F. Rouyer, Soft Matter 2013, 9, 4822. [310] A. B. Subramaniam, M. Abkarian, L. Mahadevan, H. A. Stone, Nature 2005, 438, 930. [311] M. Cui, T. Emrick, T. P. Russell, Science 2013, 342, 460. [312] X. Li, H. Shi, Y. Wang, R. Wang, S. Huang, J. Huang, X. Geng, D. Zang, Adv. Mater. Interfaces 2018, 5, 1701139. [313] K. Stratford, R. Adhikari, I. Pagonabarraga, J. C. Desplat, M. E. Cates, Science 2005, 309, 2198. [314] C. Py, P. Reverdy, L. Doppler, J. Bico, B. Roman, C. N. Baroud, Phys. Rev. Lett. 2007, 98, 156103. [315] H. Kawashima, H. Mayama, Y. Nakamura, S. Fujii, Polym. Chem. 2017, 8, 2609. [316] H. Kawashima, M. Paven, H. Mayama, H.-J. Butt, Y. Nakamura, S. Fujii, ACS Appl. Mater. Interfaces 2017, 9, 33351. [317] M. A. Nilsson, R. J. Daniello, J. P. Rothstein, J. Phys. D: Appl. Phys. 2010, 43, 045301. [318] C. Wang, G. Cui, X. Luo, Y. Xu, H. Li, S. Dai, J. Am. Chem. Soc. 2011, 133, 11916. [319] M. S. Shah, M. Tsapatsis, J. I. Siepmann, Chem. Rev. 2017, 117, 9755. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 163 [320] S. Lin, S. Nejati, C. Boo, Y. Hu, C. O. Osuji, M. Elimelech, Environ. Sci. Technol. Lett. 2014, 1, 443. [321] E. D. Bates, R. D. Mayton, I. Ntai, J. H. Davis, J. Am. Chem. Soc. 2002, 124, 926. [322] C. Wang, X. Luo, H. Luo, D. e. Jiang, H. Li, S. Dai, Angew. Chem. Int. Ed. 2011, 50, 4918. [323] M. Ramdin, T. W. de Loos, T. J. H. Vlugt, Ind. Eng. Chem. Res. 2012, 51, 8149. [324] L. D. Tijing, Y. C. Woo, J.-S. Choi, S. Lee, S.-H. Kim, H. K. Shon, J. Membr. Sci. 2015, 475, 215. 164 ACKNOWLEDGMENTS 9 ACKNOWLEDGMENTS FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 165 10 APPENDIX The following section contains important Supporting Information for Chapters 3 and 6. For Chapters 2, 4, and 5, the Supporting Information is available online; the hyperlinks are provided at the beginning of each chapter. For consistency, the numbering of the fig- ures is kept consistent with the submitted manuscripts. SUPPORTING INFORMATION FOR CHAPTER 3 Figure S2. Representative la- ser scanning confocal (LSCM) images of the nanoporous sur- faces (blue) contaminated with hydrophobic particle powder (purple) ranging between 80 nm and 50 µm before (left) and after self-cleaning (right). Some par- ticle aggregates were not di- rectly or easily cleaned, as the drops rolled over these hydro- phobic aggregates but could eventually be removed. Figure S4. Representative LSCM of the nanoporous sur- faces (blue) contaminated with hydrophilic particles from etha- nol dispersion (purple) ranging between 80 nm and 50 µm be- fore (left) and after self-cleaning (right). 166 APPENDIX Figure S5. a–c) Representative SEM images in different magnifications of the nanostructured surfaces after contamination with 600 nm hy- drophilic particles dispersed in ethanol and sub- sequent self-cleaning. Few particle aggregates could not be easily removed by sliding water drops. Figure S7. a–d) Representative SEM images in different magnifications of the nanoporous sur- faces after contamination with 200 nm (a,b) and 80 nm (c,d) hydrophilic particles dispersed in ethanol and subsequent rinsing. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 167 Figure S11. a–c) SEM images of the superhy- drophobic microstructured SU-8 pillar surface that was consecutively contaminated with powders of the hydrophobic particles (50 µm to 80 nm). After self-cleaning, 600 nm particles and especially 200 nm as well as 80 nm parti- cles remained on the surface, providing an ad- ditional roughness resulting in lower roll-off angles. Figure S12. Photographs of the superomniphobic fabrics on the car after 257 days of outdoor expo- sure. a) Superomniphobic fabrics fixed on the front window. b) Fabric fixed on the left mirror. c) Fabric fixed at the rear side win- dows. d) Fabrics fixed at the back window. All su- peromniphobic fabrics re- mained white without any obvious stains. 168 APPENDIX Figure S14. a,b) Photographs of a bare (a) and superomniphobic fabric (b) after the in- dustrial standardized contamination test us- ing a dirt dishwasher, conducted by Evonik Resource Efficiency GmbH. The bare fabric turned black due to the pickup of contamina- tion, whereas the superomniphobic remained white. c) Industrial standardized contamina- tion tests for easy-to-clean surfaces using a dirt dishwasher, conducted by Evonik Re- source Efficiency GmbH. The tested surfaces included a bare polyester fabric, superomni- phobic fabrics (SOF), and two benchmark sur- faces (BM). Table S1. Temperatures (T), rainfall, and humidities (RH) during the outdoor exposure of the superom- niphobic fabrics. Temperatures and rain total were collected from www.wetterkontor.com. The humidi- ties from www.timeanddate.de. The experiment started beginning of October 2016. Period Min. T Max. T Avg. Avg. Avg. T Rain Avg. Avg. [days] [°C] [°C] min. T max. T [°C] total min. max. [°C] [°C] [L m-2] RH [%] RH [%] 0–4 5.1 13.6 6.7 12.4 9.1 0.8 54 76 22–56 -5.4 15.3 3.0 8.4 5.7 40.6 69 93 92–133 -9.6 10.6 -1.9 3.3 0.6 32.5 66 90 189–249 -0.9 32.0 9.0 18.9 13.9 94.6 46 87 308–426 -2.8 29.1 8.3 15.5 11.8 194.7 65 96 FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 169 SUPPORTING INFORMATION FOR CHAPTER 6 Figure S1. Optical microscopic and scanning electron microscopic (SEM) images of the hexagonal plates. a,b) Optical microscopic images of the plates before (a) and after (b) fluorination treatment with 1H,1H,2H,2H-perfluorodecyltrichlorosilane. Scale bar: 0.5 mm. c,d) Scanning electron microscopic (SEM) images of the fluorinated plates at different magnifications. Scale bars: 40 µm (c), 2 µm (d). e) Side-view SEM images of the fluorinated plates. Scale bar: 40 µm. 170 APPENDIX Figure S3. Zisman plot to evaluate the critical surface energy of the fluorinated PET plates (here: conducted on a fluorinated PET substrate). The measurement should be influenced by the surface roughness of the PET substrate. Figure S4. Optical microscopic images of large polyhedral LMs. a) Polyhedral LMs pre- pared with 2.48 mm-sized plates and 120 µL, 480 µL, and 960 µL water droplets. b) Polyhe- dral LMs prepared with 0.34 mm-sized plates and 120 µL, 480 µL, and 960 µL water droplets (top). Higher magnification images are also shown (bottom). FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 171 Figure S6. SEM images of a dried polyhedral PM pre- pared with 0.34 mm-sized plates and a 3 µL water drop- let after superglue (cy- anoacrylate) treatment at different magnifications. The cyanoacrylate polymerizes and solidifies the air-water interface. The positions of the polycyanoacrylate layer as well as the air and water phase, before drying of the LM, are indicated. Figure S11. Interfacial jamming of polyhedral LMs. a) A spherocylindrical LM containing ≈100 µL of a hy- groscopic aqueous satu- rated LiCl solution (0.34 mm-sized plates) ex- posed to water vapor. The volume of the LM increased over time, leading to a de- crease in interfacial jam- ming. The LM resumed to a near-spherical shape. b) The same LM in the pres- ence of silica gel, which ab- sorbed water from the at- mosphere. The volume of the LM decreased slowly, and the LM dried, yet the LM could not return to a cylindrical shape. c) After the removal of silica gel and exposure to water vapor, the LM resumed to the near-spherical shape again due to water uptake. Scale bars: 2 mm. 172 APPENDIX Figure S12. NH3-induced color change of polyhedral LMs. a) The ammonia gas penetrated through the gaps between the plates of a 30 µL polyhedral LM prepared us- ing 1.13 mm-sized transparent plates. Therefore, a slight color gradient is visible between the edges and the middle of each plate in the aqueous solution un- derneath. Figure S13. THF-induced breakage of a polyhedral LM. a) Stimuli-re- sponsive 30 µL polyhedral LM (0.34 mm-sized plates) containing dyed water broke after 5–6 min ex- posure to THF vapor. b) Backside view of broken liquid marble. The liquid marble broke on the bottom right edge (marked with white dots), and the liquid spread over the glass slide afterward. Scale bars: 2 mm. Figure S14. Broken polyhedral LM (containing ≈60 µL aqueous saturated LiCl solution) floating on a water surface. The magni- fying effect of the water droplet on top of the plates allows close observation of the plates sepa- rating the two air-water inter- faces. FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 173 11 CURRICULUM VITAE 174 CURRICULUM VITAE FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 175 12 LIST OF PUBLICATIONS AND CONTRIBUTIONS PUBLICATIONS 1. C. Shi, X. Cui, X. Zhang, P. Tchoukov, Q. Liu, N. Encinas, M. Paven, F. Geyer et al., Interaction between air bubbles and super-hydrophobic surfaces in aque- ous solutions, Langmuir 2015, 31, 7317. 2. F. Geyer, C. Schönecker, H.-J. Butt, D. Vollmer, Enhancing CO2 Capture Using Robust Superomniphobic Membranes, Adv. Mater. 2017, 29, 1603524. 3. K. Jayaramulu, F. Geyer, M. Petr, R. Zboril, D. Vollmer, R. A. Fischer, Shape Controlled Hierarchical Porous Hydrophobic/Oleophilic Metal-Organic Nano- fibrous Gel Composites for Oil Adsorption, Adv. Mater. 2017, 29, 1605307. 4. N. Gao, F. Geyer, D. W. Pilat, S. Wooh, D. Vollmer, H.-J. Butt, R. Berger, How drops start sliding over solid surfaces, Nat. Phys. 2018, 14, 191. 5. H. Teisala, F. Geyer, J. Haapanen, P. Juuti, J. M. Mäkelä, D. Vollmer, H.-J. Butt, Ultra-fast processing of hierarchical nanotexture for a transparent superam- phiphobic coating with extremely low roll-off angle and high impalement pressure, Adv. Mater. 2018, 30, 1706529. 6. F. Geyer, M. D’Acunzi, C.-Y. Yang, M. Müller, P. Baumli, N. Encinas, D. Vollmer, H.-J. Butt, How to Coat the Inside of Narrow and Long Tubes with a Super-Liquid-Repellent Layer – A Promising Candidate for Antibacterial Cath- eters, Adv. Mater. 2019, 31, 1801324. 7. F. Geyer, P. M. Seiler, D. Vollmer, R. Berger, Wie sich Tropfen bewegen, Phys. Unserer Zeit 2019, 50, 90. 8. P. Baumli, H. Teisala, H. Bauer, D. Garciá-González, V. Damle, F. Geyer et al., Flow-induced long-term stable slippery surfaces, Adv. Sci. 2019, 1900019. 9. F. Geyer, Y. Asaumi, D. Vollmer, H.-J. Butt, Y. Nakamura, S. Fujii, Polyhedral Liquid Marbles, Adv. Funct. Mater. 2019, 1808826. 10. F. Geyer, M. D’Acunzi, A. Sharifi-Aghili, A. Saal, N. Gao, A. Kaltbeitzel, D. Vollmer, H.-J. Butt, When and how self-cleaning of superhydrophobic sur- faces works, 2019, submitted. 176 LIST OF PUBLICATIONS AND CONTRIBUTIONS 11. C.-Y. Yang, N. Encinas, F. Geyer, A. Kaltbeitzel, P. Baumli, J. Reinholz, V. Mailänder, D. Vollmer, H.-J. Butt, Sub-micrometer sized roughness sup- presses bacteria adhesion, 2019, submitted. PATENT APPLICATIONS 1. H.-J. Butt, D. Vollmer, F. Geyer, C.-Y. Yang, N. Encinas, M. D’Acunzi, EP17207768.7 patent application, 2018. PRESENTATIONS (SELECTED) FUNDAMENTALS AND APPLICATIONS OF SUPER-LIQUID-REPELLENT SURFACES 177