Delaying Frost Formation by Controlling Surface Chemistry of Carbon

Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)

Article

Delaying Frost Formation by Controlling Surface Chemistry of Carbon Nanotube-Coated Steel Surfaces Yu Zhang, Mena Klittich, Min Gao, and Ali Dhinojwala ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.6b11531 • Publication Date (Web): 24 Jan 2017 Downloaded from http://pubs.acs.org on January 25, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Delaying Frost Formation by Controlling Surface Chemistry of Carbon Nanotube-Coated Steel Surfaces Yu Zhang,†,¶ Mena R. Klittich,†,¶ Min Gao,‡ and Ali Dhinojwala∗,† †Department of Polymer Science, The University of Akron, Akron, Ohio 44325-3909, USA ‡Liquid Crystal Institute, Kent State University, Kent, Ohio 44242, USA ¶These two authors contributed equally. E-mail: [email protected] Abstract Superhydrophobic surfaces are appealing as anti-icing surfaces, given their excellent water repellent performance. However, when water condenses on the surface due to high humidity, the water becomes pinned, and superhydrophobic surfaces fail to perform. Here we studied how the stability of the superhydrophobicity affected water condensation and frost formation. We created rough surfaces with the same surface structure, but with a variety of surface chemistries, and compared their anti-frost properties as a function of intrinsic contact angle. Frost initiation was significantly delayed on surfaces with higher intrinsic contact angles. We couple these macro measurements with environmental scanning electron microscopy of water droplet initiation under high humidity conditions. These provide experimental evidence towards previous hypotheses that for a lower intrinsic angle rough substrate, Wenzel state is thermodynamically favorable, while the higher intrinsic angle surface maintains a Cassie-Baxter state. Surfaces with a thermodynamically stable Cassie-Baxter state can then act both as

1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

anti-steam and anti-frost surfaces. This research could answer the persistent question of why superhydrophobic surfaces sometimes are not icephobic; anti-icing performance depends on the surface chemistry, which plays a critical role in the stability of the superhydrophobic surfaces.

Keywords hydrophobicity, superhydrophobic, surface energy, surface modification, water, ice, freezing

1

Introduction

Ice formation on the surfaces of turbine blades, aircrafts, antennas, boats, or electronic devices has aroused concerns of safety, inefficiency, and energy costs. 1,2 To solve these problems, considerable literature in this field focused on developing anti-icing surfaces or coatings. Among these studies, superhydrophobic surfaces have been reported as an effective way to delay the freezing time of water droplets. 3–6 However, the mechanisms of why superhydrophobic surfaces can delay the ice freezing time are not clear. To date, several factors, such as the heat transfer between substrate and water, surface energy, surface roughness, nucleus curvature, water droplet size, surface area, heat transfer, surface epitaxial matching, and the interfacial energy barrier 4–13 have been attributed to delay the freezing of water droplets at superhydrophobic surfaces. Based on the unclear mechanism, the role of surface roughness of superhydrophobic surfaces on the anti-icing behavior has been debated, 14–16 and literature has emerged that offers contradictory findings about superhydrophobic surfaces, namely that they are not always icephobic. 17–21 Some superhydrophobic surfaces will fail to inhibit icing when exposed to a humid environment, resulting in water condensation or frost formation. 22 Most superhydrophobic surfaces are not anti-fog or anti-frost, and once fog or frost has formed on their surface, the surface chemistry is changed, becoming superhydrophilic. This shift results in 2 ACS Paragon Plus Environment

Page 2 of 24

Page 3 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


the initially anti-icing surface becoming an ice promoting surface. Additionally, ice adhesion strength on an initially superhydrophilic surface and a superhydrophobic one is found to be almost the same, as the mechanical interlocking between the ice and the surface texture results in strong adhesion. 23 Thus the superhydrophobic icephobic surface transforms into a superhydrophilic surface with strong ice adhesion as a result of condensation. To create a superhydrophobic surface which could have a robust anti-frost or anti-icing property in high humidity conditions, researchers have used slippery liquid-infused porous surfaces (SLIPS) to create a low adhesion surface. 2,24 Once ice forms, it easily slides off of the cold surface due to the low friction between the ice and the substrate. Compared to the frost coverage for bare aluminum, greased aluminum, and superhydrophobic coated aluminum surfaces (when tilted at a 70◦ angle), the SLIPS-coated substrate had the lowest frost coverage. A challenge to use these liquid-infused porous structures is their limited durability, as the liquid diminishes over time. 25 We try to solve the problem from a different angle: frost prevention. Frost is a critical indicator of the surface transitioning from superhydrophobic to hydrophilic under cold humid conditions. Understanding the condensation formation and removal from rough surfaces would pave the way for development of robust anti-icing surfaces. Our previous study has found that a carbon nanotube based surface plasma coated with a fluorinated polymer (PFCNT) can withstand thermal shocks and has a robust anti-steam property. It remains a dry structure in humid conditions for at least 10 hours, whereas the pristine carbon nanotube based surface (CNT) will wet (contact angle decreases to 0◦ ) within a 10 minute exposure to steam. 26 Given that these two surfaces have similar surface roughness, area, and heat transfer, it is striking how differently they perform; only their surface chemistries differ. By applying thermodynamic calculations, the hypothesis for this phenomenon is that the PF-CNT is thermodynamically stable at the Cassie-Baxter state, while the uncoated-CNT is thermodynamically favorable of Wenzel state. 26 In the Cassie-Baxter state, a droplet of water on a surface will be supported on the air-solid composite surface, whereas in a Wenzel state,



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

water completely fills the roughness asperities on the surface. 27 Therefore a surface that is stable in the Cassie-Baxter state should show both anti-steam and anti-frost performance under high humidity. Here, we explore the role of surface chemistry further, investigating frost coverage on coated CNT surfaces with varying wetting properties. Previous work has tracked the contact angle of a surface impinging droplet over time below 0 ◦ C; it has been found that on differently rough but chemically similar surfaces, the contact angle on hydrophobic surfaces decreases as a function of temperature, due to water condensation (when under 40 % relative humidity). Under this higher humidity, the expected ice delay for the superhydrophobic surface was lost, indicating that the stability of the Cassie-Baxter state on the superhydrophobic surface may then be affected by temperature, contributing to poor anti-icing performance in high humidity. 15 On a fully fluorinated rough surface, Feng et al. found that the dynamic contact angle dropped only ∼2◦ between 0 and -10 ◦ C at 60 and 90 %RH, indicating that on a surface with sufficient roughness, structure alignment, and high intrinsic contact angle, the Cassie-Baxter state may be stable even under high humidities. 28 Here, we have used our carbon nanotubes to create a rough, randomly oriented substrate, with varying intrinsic contact angles. We used plasma enhanced chemical vapor deposition (PECVD) to change the surface chemistry of our CNT substrate. The PECVD process enables the deposition of a very thin layer of a film onto the CNT without disturbing its structural integrity, allowing comparison of surface chemistries while maintaining a constant roughness. 26 The CNT structures were coated to give a range of intrinsic contact angles from 40◦ to 110◦ , where the PF-CNT is the most hydrophobic. Since these surfaces undergo condensation within a rough porous membrane, it is difficult to experimentally verify the wetting state. We studied the water condensation mechanism by experimentally monitoring the water condensation process on PF-CNT, compared to uncoated CNT substrates, in an oversaturated humid environment using environmental scanning electron microscopy (ESEM). Combining the ESEM with the frost coverage exper-


Page 4 of 24

Page 5 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


iments provides both micro and macro scale observations of surface’s anti-frost performance. In this way, we could understand how surface stability and surface chemistry contribute to the anti-frost/anti-steam performance of superhydrophobic surfaces, providing a novel way to design anti-icing surfaces.

2

Experimental Section

2.1

Sample preparation

Carbon nanotube surfaces were made using a floating catalyst deposition method, where xylene was the carbon source and ferrocene was used as the catalyst in a chemical vapor deposition furnace (Nanotech Innovations LLC, Oberlin OH). The catalyst (1 gram ferrocene) was dissolved into the carbon source (100 mL xylene), and injected into the furnace at a rate of 0.11 mL/min. A total of 9.8 mL of the xylene/ferrocene mixture was injected, passing through a 250 ◦ C pre-heat zone before reaching the growth zone, set to 760

◦

C. Argon and

hydrogen were used as carrier gases (85:15 v/v). Following the procedure outlined in Sethi et al., 26,29 the CNTs were grown on 1 × 1 cm acid etched stainless steel (SS 304) plates, resulting in highly porous rough CNT structures. These CNT tubes are decorated with amorphous carbon, increasing their surface roughness on the nanoscale (representative scanning electron microscopy (SEM) images are included in the Supplementary Information). Following the same technique as outlined in Badge et al., 2011, the surface chemistry of the CNT surfaces were altered using radio frequency plasma-enhanced chemical vapor deposition (PECVD). 26 Using PECVD, thin layers of 1H,1H,2H-perfluoro-1-dodecene (PF), maleic anhydride (MA), and hexamethyldisiloxane (HMDSO) were coated onto CNT surfaces, creating surfaces with 3 different surface chemistries but the same roughness characteristics. Ignition vapor pressures for PF, MA, and HMDSO were 200, 50, and 200 mTorr, respectively. Both CNT samples and silicon wafer references were coated during the same deposition. Depositions were done by allowing the precursor to vaporize for 5 minutes, fol5 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

lowed by a continuous 5-minute plasma deposition using an input power of 35 Watts. After the plasma was turned off, the samples were exposed to the precursor vapor for an additional 5 minutes to react any remaining radicals. Previous work with these precursors and deposition protocol showed that plasma coated MA surfaces exhibit strong C1 and O1 peaks when characterized using X-ray Photoelectron Spectroscopy (XPS). 30 Plasma coated HMDSO surfaces have peaks corresponding to Si-CH3 or Si-CH2 -CH2 - bonds, as well as -C-O- bonds. 31 XPS was also used to compare uncoated CNTs to PF-CNTs; PF has characteristic C1 and F1 peaks, corresponding to C-Fn bonds. The measured PF-CNT also exhibited C-C bonds, a characteristic bond from the underlying CNTs. 26 The authors concluded that the presence of this peak (together with their measurements from transmission electron microscopy) indicated that the PF coating was less than 10 nm thick (the probe depth of the XPS). The same experimental procedures was followed for both CNT growth and plasma coating.

2.2

Contact angle

Contact angles of both the coated reference silicon wafers and the coated CNTs were measured using a Ram´e-Hart Instruments Advanced Goniometer. A droplet of 10–12 µL of ultra-pure water (18.2 MΩ/cm) was deposited on the surface and the static contact angle was measured after 5 seconds. To compare the hydrophobicity of the surface chemistry, we used the contact angle (CA) to characterize the surface coatings on the coated reference silicon wafers (Table 1). We also measured the contact angles of pendant droplets on MA, HMDSO, and PF coated CNT, as well as the uncoated CNT substrates. The MA-CNT had a contact angle of ∼0◦ ; when the droplet touched to the MA-CNT surface, it was quickly absorbed by the substrate. However, the uncoated CNT showed an initial contact angle of ∼151 ± 3◦ , which decreased with time. After 2-3 minutes, the contact angle decreased to 0◦ . The contact hysteresis for the CNT at 5 seconds was ∼5◦ . The HMDSO-CNT and PF-CNT had a stable contact angles of 150 ± 3◦ and 165 ± 1◦ respectively, with contact angle hys6 ACS Paragon Plus Environment

Page 6 of 24

Page 7 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


teresis of ∼0◦ . They maintained these stable contact angles after frost testing (values and images are provided in Supporting Information section 2). Table 1: Coating characterization: contact angles for coated CNT substrates, and coating thicknesses on the silicon wafer controls. Values are means, ±1 standard error. Contact Angle Surface CNT Coated Silicon Wafer ◦ ◦ Uncoated 145 ± 9 N/A MA wetting 44◦ ± 2.4◦ HMDSO 150◦ ± 3◦ 91◦ ± 0.3◦ PF 165◦ ± 1◦ 105◦ ± 4◦

2.3

Frost coverage

The anti-frost performance of the MA-CNT, CNT, HMDSO-CNT, and PF-CNT substrates was characterized on a heating and cooling stage (MK1000 series from Instec Inc.) in a walk-in humidity chamber with a stable room temperature of 25 ± 1 ◦ C and a controllable percent relative humidity (% RH) of 60 ± 5. All four surfaces were placed on the same stage, allowing for direct comparison between substrates. Pieces of double sided copper tape were applied between the cooling stage and the samples to improve contact between the stage and the substrates. The substates were placed on the level stage at room temperature (25 ± 0.5 ◦ C) and then cooled to 5 ◦ C at a rate of 5 ◦ C/min. A timer was started, and the stage was then brought to -2 ◦ C using 2 ◦ C/min, then held at -2 ◦ C for the duration of the test. A digital camera (Olympus OM-D E10) was positioned directly above the stage using a tripod. Photographs were taken every 1 minute for the first 20 minutes, then every 10 minutes thereafter (photographs from one trial set are compiled in Figure 1). Frost coverage was analyzed using ImageJ; edge effects were included. Three sets of frost coverage trials were done; average frost/ice coverage ± standard error were calculated (Figure 2).



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

2.4

Environmental scanning electron microscopy

The environmental scanning electron microscopy (ESEM) observations were carried out in a FEI Quanta 450 environmental SEM. A Peltier cooling stage was used to control the sample temperature. The surface was maintained at a temperature of 15 ◦ C, with humidity in the oversaturation regime (using water vapor pressure of 15 Torr). We recorded ESEM videos for both PF-CNT and uncoated-CNT, which are provided in supporting information. The water condensation experiments by ESEM have been repeated 15 times at different locations, with the same phenomena observed at each location.

3 3.1

Results and Discussion Frost performance

The frost formation coverage over time for the four samples is presented in Figure 2. The rate of frost coverage is temperature dependent; 32 at low temperatures, such as -10 ◦ C, the differences between the different coated surfaces were indistinguishable using our set-up. Using the temperature of -2 ◦ C enabled us to clearly distinguish the frost formation on a macro scale (Figure 1). As the surface was cooled to -2 ◦ C, condensation was observed on HMDSO-CNT and PF-CNT, but neither the MA-CNT nor the uncoated CNT had distinct drops, appearing wet. Frost initiated on MA-CNT first, followed closely by CNT, and then HMDSO-CNT. Droplets were observed to roll off of the surface of both HMDSO-CNT and PF-CNT, leaving clean dark ‘trails’ behind them. These droplets kept merging and rolling off of the surface until we stopped the experiments (as long as 120 min), indicating the surfaces had reached an equilibrium state of water condensing and self-removing. As the frost experiment continued, frost formed on the steel stage; as the droplets rolled off of the hydrophobic substrates, they froze on the frosted stage, gradually encroaching on the sample surface and covering both the HMDSO-CNT and PF-CNT with ice. Due to the small size


Page 8 of 24

Page 9 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


of the samples (1 cm x 1 cm), these edge effects were quite significant over time (in Figure 2, ice coverage is indicated with solid symbols). While some frost was seen on the HMDSOCNT, it was quickly engulfed in the advancing edge ice. For the PF-CNT, there was no frost formation on the surface of the substrates, however there was a small amount of frost growing on the vertical edge along the steel surface of the cooling stage. The advance of frost and ice formation from the edge has been widely studied in literature; Boreyko and Collier have quantified the advancement rate of frost from the edges, a phenomenon which they attribute to interdrop freezing. 33 On superhydrophobic surfaces, supercooled water condensates have been shown to self propel off of the surface before heterogeneous ice nucleation occurs; 1,28,32,33 this phenomenon can affect the rate of condensation frosting, decreasing it by 3 times over surfaces without jumping condensates. 33 We observed jumping condensates on both the HMDSO-CNT and the PF-CNT, but were not able to quantify it with our set-up. It is important to note that, as we used the PECVD to deposit our surface coating, the edges of the HMDSO-CNT and the PF-CNT were also hydrophobic. As the drops moved, both through jumping and rolling, they would encounter the leading edge of the advancing ice (originating from the hydrophilic steel support surface). This edge effect dominated the frost formation on the horizontally placed hydrophobic samples. Had our substrates been at an angle to the surface (or even upside down), 32 the jumping condensates could have been removed during the frosting process, reducing the rate of frost formation. Comparing the HMDSO-CNT and PF-CNT, we observed that the HMDSOCNT tended to frost more quickly (Figure 1), although the total average surface coverage of HMDSO-CNT was comparable to the total surface coverage of PF-CNT (Figure 2). For our MA-CNT and uncoated CNT, we saw frost initiation across the surface of the sample (See Figure 1, 11 minute time point), rather than from the edge. Frost nucleation on MA-CNT and uncoated CNT was heterogeneous, correlating with their Wenzel wetting state. We have also studied the frost formation when the temperature was decreased to -10 ◦ C; both uncoated CNT and PF-CNT surfaces (placed horizontally) were covered with ice within 6



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

minutes. Because at -10 ◦ C we observed freezing of discrete macro droplets (as a result of coalescence) rather than frost formation, we report here only the frosting comparison at -2 ◦ C. The superhydrophobic surfaces (HMDSO-CNT and PF-CNT) are also expected to delay the freezing of macroscopic water droplets due to the decrease in contact surface area available for surface nucleation of ice. 4


Page 10 of 24

Page 11 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


MA-CNT

CNT

HMDSO-CNT

PF-CNT

0 min

11 min

20 min

40 min

60 min

75 min

90 min

120 min

Figure 1: Image of the surfaces covered by ice as a function of time. The substrate temperature was -2 ± 0.1 ◦ C. The air temperature was 25 ± 0.5 ◦ C. The RH was 60 ± 5 %. The four kinds of substrates were placed at the same cooling stage with a environmental temperature at 25 ◦ C, and cooled at the fixed rate (at 2 ◦ C/min) from 5 ◦ C to -2 ◦ C simultaneously, and then maintained at -2 ◦ C. Time was measured from when the temperature stage reached -2 ◦ C. The ice-covered area is shown as white, and the area without ice is shown as black. After 20 min of freezing, >95 % of MA-CNT surfaces are covered with ice. After 40 min of freezing, >95 % of CNT surfaces are covered with ice except for superhydrophobic samples (HMDSO-CNT and PF-CNT). After 60 min of freezing, >20 % of HMDSO surfaces are covered with ice while only ∼5 % of PF-CNT surface is covered with ice, growing from the frost of the cooling stage. Some ice formation was observed on PF-CNT, which we attribute to the edge effect from the frost growing from the cold plate; the bulk of PF-CNT remains ice-free. The substrates are placed horizontally without tilting and are approximately 1 cm × 1 cm in size. See Figure 2 for the characterization of the surface area covered by ice as a function of cooling time. 11 ACS Paragon Plus Environment


100

CNT

80

MA-CNT HDMSO-CNT

)%( egarevoC ecI/tsorF

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 24

PF-CNT

60

40

20

0 0

20

40

60

80

100

120

o

Time at -2 C (minutes)

Figure 2: Area fraction of the surfaces covered by frost (open symbols) and ice (solid symbols) as a function of time on the horizontally placed 1 cm× 1 cm size substrates. The four substrates (MA-CNT, CNT, HMDSO-CNT, PF-CNT) were placed on the same cooling stage with a environmental humidity of 60 % and temperature at 25 ◦ C, cooled at a fixed rate (at 2 ◦ C/min) from 5 ◦ C to -2 ◦ C simultaneously, then maintained at -2 ◦ C. The error bars represent ± standard error from three individual measurements of frost formation. The ice coverage of the PF-CNT stems from ice at the substrate’s edge, which is growing from the cooling stage, and the bulk of PF-CNT remains ice-free. The images were shown in Figure 1.

After the melting process, when we increased the substrate temperature to room temperature, the PF-CNT returned back to a dry structure whereas HMDSO-CNT retained a few droplets, which were easily removed when gently blown. Contact angles of both the PF-CNT and the HMDSO-CNT were taken after frost trials were performed, and are presented in the Supporting Information Table S1. There was no decrease in contact angle after the frost cycle. In contrast, the hydrophilic substrates, CNT and MA-CNT, were totally wetted by a thin layer of water film which resulted from the melting of the frost layer. From the contact angle measurements, we found that the MA-CNT is superhydrophilic. It absorbed the water droplet in less than 1 second. The uncoated CNT shows an initial contact 12 ACS Paragon Plus Environment

Page 13 of 24

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


angle around 151◦ , then the contact angle decreases with exposure time. The HMDSO-CNT and PF-CNT present stable contact angles of 150◦ and 165◦ respectively, and water droplets will remove easily off of the surface due to the small contact angle hysteresis and the surface roughness. From literature, the intrinsic contact angle of the surface of a single CNT has been measured to be about 80◦ . 34 Here, the degree of hydrophobicity could be considered as: MA-CNT

Delaying Frost Formation by Controlling Surface Chemistry of Carbon

Recommend Documents