Silicone-Infused Antismudge Nanocoatings - ACS Applied Materials

Feb 22, 2017 - aAfter sitting under a water droplet for a long time, the surface PDMS chains may retreat into their hosting nanopools as shown to the ...
9 downloads 22 Views 6MB Size
Research Article www.acsami.org

Silicone-Infused Antismudge Nanocoatings Heng Hu, Jian Wang, Yu Wang, Emily Gee, and Guojun Liu* Department of Chemistry, Queen’s University, 90 Bader Lane, Kingston, Ontario, Canada K7L 3N6 S Supporting Information *

ABSTRACT: A polyurethane-based NP-GLIDE coating that bears on its surface and in its interior nano-pools of a grafted liquid ingredient for dewetting enablement is obtained from casting and curing a film comprising a triisocyanate, a polyol (P1), and a graft (g) copolymer of P1 and poly(dimethylsiloxane) (P1-g-PDMS). A silicone-infused NP-GLIDE (SINP-GLIDE) PU coating is obtained from cocasting the NP-GLIDE precursors with a free silicone oil (SO) or SO mixture (SOs). This paper reports the preparation of the novel SINP-GLIDE coatings and discusses the effect of changing the amount and type of the infused SO as well as the coating formation conditions on their optical clarity. Also reported are the contact and sliding angles of various test liquids on the NP-GLIDE and SINP-GLIDE coatings, and the data variation trends are rationalized using existing theories. Further, the stable water sliding performance of the SINP-GLIDE coatings under simulated raining and other conditions is demonstrated. The improved and stable water sliding performance of the SINP-GLIDE coatings facilitates their practical applications. KEYWORDS: coatings, antismudge, slippery surfaces, polyurethane, antigraffiti

I. INTRODUCTION An NP-GLIDE coating bears on its surface and in its interior nanopools of a grafted liquid ingredient for dewetting enablement (Scheme 1a). The preparation of both polyur-

copolymer and cannot macrophase separate from other components. Because the nanopools do not scatter much light, the NP-GLIDE coatings are optically clear. On these surfaces even at a PDMS bulk content of as little as 4.0 wt %, ink, paints, and simulated fingerprints readily shrink because PDMS self-enriches on the coating’s surface and converts a solid PU surface into a lubricated liquidlike one.1−3 Further, droplets (5 μL) of most solvents with surface tensions >23 mN/m cleanly glide down at substrate tilt angles of less than 5° without leaving a trace. Moreover, such a coating is weartolerant and retains its antismudge properties after uniform wearing that evenly decreases the coating thickness without extensive surface roughening because the coating surfaces are self-regenerating: when the surface layer is worn away, nanopools that were initially embedded underneath are ruptured, and chains of the freshly released liquid polymer replenish the worn surface. Despite the many desirable features of NP-GLIDE PU coatings and their practical value, they have mediocre water sliding properties. For example, 10 μL water droplets could not slide on them. At a droplet volume of 15 μL, substrate tilting angles of higher than 30° were required to enable water sliding.1−3 This mediocre water sliding performance hinders their applications in general and in particular their use on automobile windshields, for example, where the ready shedding of rain droplets renders the deriver improved visibility. This paper reports the infusion of some silicone oil (SO) into the NP-GLIDE coatings to yield SO-infused NP-GLIDE or SINP-

Scheme 1. (a) Schematic Cross-Sectional Structure of a NPGLIDE Coating Containing PDMS as the Dewetting Enabler;a (b) Schematic of the Ideal Cross-Sectional Structure for a SO-Infused NP-GLIDE Coating

a

After sitting under a water droplet for a long time, the surface PDMS chains may retreat into their hosting nanopools as shown to the left. In air, the PDMS chains may escape from the nanopools to cover the coating’s surface as shown in the middle. Under a hexadecane droplet, the escaped chains become swollen by the test solvent as shown to the right. The liquid droplets and nanopools are not drawn to scale.

ethane-based (PU-based)1,2 and epoxy-based3 NP-GLIDE coatings has been reported by us. To prepare a PU-based NP-GLIDE coating, we cast a solution containing a triisocyanate, a polyol oligomer (P1), and a graft (g) copolymer (P1-g-PDMS) consisting of poly(dimethylsiloxane) grafted onto P1, evaporate the solvent, and then cure the formed film. Here the dewetting enabler PDMS is dispersed eventually as nanopools in the coating because it is introduced as a graft © XXXX American Chemical Society

Received: January 4, 2017 Accepted: February 22, 2017 Published: February 22, 2017 A

DOI: 10.1021/acsami.7b00126 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces GLIDE coatings. Under optimized conditions, the new coatings retain essentially all of the advantageous features, including the optical clarity, of the NP-GLIDE coatings but gain much improved water sliding performance. This paper also reports the contact and sliding angles of various organic liquids on the NP-GLIDE and SINP-GLIDE coatings and justifies the observed data variation trends. It further demonstrates the stability of the water sliding performance of the SINP-GLIDE coatings under simulated raining and other conditions and thus their practical value. Antismudge coatings reported in the literature are still mostly based on low-surface-tension textured rough surfaces,4−18 despite that this approach yields coatings that are opaque and not wear resistant.19 Our NP-DLIDE and SINP-GLIDE coatings were inspired by prior reports of grafted liquidlike monolayers20−23 and of slippery liquid-infused porous surfaces (SLIPS).24−31 In the former case, a monolayer of poly(dimethylsiloxane) (PDMS)20−22 or highly mobile organic molecules23 are grafted onto flat solid substrates to render the dewetting ability. NP-GLIDE coatings have improved over these monolayer coatings because of their enhanced wear tolerance and ability to bind to many substrates. In the case of a SLIPS, the pores of a premade porous surface are filled with a low-surface-tension liquid.24−32 A test liquid readily slides down off the surface of the first liquid held in position by the porous substrate because the two liquids are immiscible and the friction coefficient between the two liquids is low. Our SINP-GLIDE differs from a SLIPS because of the scaffolds used for hosting SO is a structurally unique NP-GLIDE coating. Despite its unique internal structure, our SINP-GLIDE coating is a lubricant-infused cross-linked coating. The latter has already found or is promising applications. For example, SO has been incorporated into cross-linked PDMS coatings to reduce the adhesion of marine fouling organisms.33−36 More recently, SO-infused PDMS coatings have been found to reduce the deposition of bacteria,37 ice,30 dust particles,29 or water.38 Aside from cross-linked PDMS, a PDMS-containing polyurea has also been used as the host for SO.36 Even a multilayer coating from the layer-by-layer deposition has been used as the host for a fluorinated liquid lubricant.31 We reasoned that SO incorporation into an NP-GLIDE coating would improve its water sliding performance because SO would help increase the coverage of the coating surface by PDMS (Scheme 1b). To prepare NP-GLIDE, PDMS was introduced as P1-g-PDMS. While the use of P1-g-PDMS offered many of the reported properties, the polyol backbone competed with PDMS for surface sites and limited the surface density of PDMS. On the surface of a NP-GLIDE coating, PDMS formed discontinuous nanopools rather than a continuous film (Scheme 1a). Infused silicone oil (SO) would increase the surface coverage by PDMS because the SO chains do not bear a tethering backbone (Scheme 1b) and may form a continuous film on the coating surface together with the grafted PDMS chains.

Scheme 2. Chemical Structures of (a) HDIT and (b) P1-gPDMS

ratios of 23.9%, 20.6%, 24.0%, and 31.5%, respectively, were copolymerized via free radical polymerization. These feed ratios were similar to those of a commercial polyol that we used in a previous investigation.2 To simulate the conditions likely used in industry, we completely polymerized the monomers. Thus, the resultant polymer should have an overall chemical composition dictated by the monomer feed molar ratios (Scheme 2b). However, the polymer should have a high degree of composition heterogeneity due to polymer composition drifts over the course of the polymerization. The latter arose due to the high monomer conversion and the different reactivity ratios of the monomers. Our size exclusion chromatography (SEC) analysis based on PS standards gave a number-average molecular weight Mn of 9.7 × 103 Da and a dispersity index of DI = 1.60 for the P1. On the basis of this PS-equivalent Mn value, we calculated a repeat unit number of 80 for P1. The absolute molecular weight of P1 was not determined because the coating properties should not significantly change with a slight variation in the molecular weight of P1. In addition, the absolute molecular weight was not needed for calculating formulation parameters such as the final PDMS weight fraction in the coating or the feed mass ratios of the different components. P1-g-PDMS Synthesis and Characterization. To prepare P1-g-PDMS, P1 was reacted with PDMS-COCl, which denoted PDMS chains bearing a terminal acid chloride group and possessing a molecular weight of 4.7 × 103 Da.1,2 Because of the relatively high molecular weight of P1 and the low grafting density of PDMS, the resultant P1-g-PDMS did not readily disperse in hexanes to form micelles. Thus, we extracted the reacted mixture with hexanes to selectively remove ungrafted PDMS-COCl or P1-g-PDMS chains that were rich in PDMS and formed micelles in hexanes. After it had been extracted six times with hexanes, the purified P1-g-PDMS contained 31 wt % PDMS as determined from 1H NMR. While this value was lower than 42 wt % for the unextracted sample, it did not decrease further after the sample that was extracted six times were additionally extracted twice with hexanes. Thus, the purified P1-g-PDMS was free of ungrafted PDMS. Using the weight fraction of 31%, the molecular weight of 4.7 × 103 Da for PDMS, and the repeat unit number of 80, we calculated the molar fractions of all the components in P1-g-PDMS. Scheme 2b shows the structure of P1-g-PDMS thus determined. On the average, each P1 chain bore 0.96 PDMS chains. Coating Preparation. The coating samples were prepared by drop-casting solutions onto 1.0 × 1.0 in.2 glass slides,

II. RESULTS AND DISCUSSION P1 Synthesis and Characterization. Our NP-GLIDE coating was prepared by casting a solution comprising a polyol (P1), P1-g-PDMS, and a hexamethylene diisocyanate trimer (HDIT, Scheme 2a) onto a substrate and then curing the coating at 120 °C.1,2 To prepare the targeted P1, styrene (S), methyl methacrylate (MMA), 2-hydroxylethyl methacrylate (HEMA), and butyl methacrylate (BMA) at the feed molar B

DOI: 10.1021/acsami.7b00126 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 1. AFM topography images of cross sections of (a) PU-3.0%, (b) PU-6.0%, (c) PU-9.0%, and (d) PU-12.0% coatings.

Figure 2. Transmittance curves of various coatings: (a) PU-3.0%-y% and (b) PU-6.0%-y% coatings infused with different amounts of SO-1, SO-2, and SO-3, respectively. (c) PU-6.0%-y% coatings infused with SO-3 cast at room temperature and 45 ± 5 °C, respectively. (d) SO-1-containing PU6.0%-y% coatings cast at room temperature and PU-6.0%-y% coatings, cast at 45 ± 5 °C, containing a mixture of SO-1, SO-2, and SO-3 at the mass ratios of 2.00/2.00/1.00.

evaporating the solvent, and then curing the dried film at 120 °C for 12 h. To produce a PU-x% or NP-GLIDE coating that contained x wt % of grafted silicone (GS) in the final coating, a

solution containing P1, P1-g-PDMS, and HDIT was cast onto a glass slide. The x% value was tuned by adjusting the relative mass ratio between P1 and P1-g-PDMS. We avoided the C

DOI: 10.1021/acsami.7b00126 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 3. AFM height images showing cross sections of (a) PU-6.0%-16.0% and (b) PU-6.0%-32% coatings.

coating samples and to decrease the system’s surface energy (Scheme 1). The coatings were highly clear because the average diameter of the wells never exceeded 22 nm, which was much smaller than the wavelength of the visible light. Effect of SO Incorporation on the Clarity of the SINPGLIDE Coatings. The three types of SOs used were SO-1, SO2, and SO-3 and possessed kinematic viscosities of 5, 20, and 50 cSt as well as molecular weights of 0.77 × 103, 2.0 × 103, and 3.8 × 103 Da, respectively. To quantify the clarity of the SOcontaining coatings with a controlled thickness of 30 μm, we measured the transmittances at 500 nm of three samples of each coating and in some 10 regions for each sample and then calculated the average transmittance and standard deviation. Figures 2a and 2b show the variations in the transmittance T% of the PU-3.0%-y% and PU-6.0%-y% samples as a function of y %. As the SO weight percentage y% among SO and GS increased, transmittance T% decreased. The T% value decreased more sharply with y% as the molecular weight of the SO increased. For SO-1-containing coatings, T% barely changed with y% below y% = 8%. In every case, the T% decrease was accompanied by greater fluctuations in the measured T% values. The T% value declined more sharply with y% as the molecular weight of the SO increased because the compatibility between SO and GS decreased as the molecular weight of the SO increased and approached 4.7 × 103 Da for the GS. Previous studies of polymer blends consisting of a homopolymer A and an AB diblock copolymer revealed that the homopolymer A would mix with the A block or swell the A domains of the diblock copolymer only if the molecular weight of the A polymer was substantially lower than that of the A block.40 As the molecular weight of the A polymer approached that of the A block, the A block domain was no longer swollen by the A polymer anymore, and the system entered the dry brush regime. In the dry brush regime, macrophase separation between the A polymer and the A block occurred. In our case, an enhanced macrophase separation between SO and GS as the SO molecular weight increased caused T% to decrease more sharply with y%. The T% value did not decrease significantly with y% below 8% for the SO-1-containing coatings probably because the SO was first dissolved into the GS nanopools. It was only after the uptake capacity of the nanopools was saturated (SO uptake enhances GS chain stretching) that the added SO underwent macrophase separation and the coating’s clarity sharply decreased. Macrophase separation also intensified fluctuations

preparation of SO-infused NP-GLIDE coatings by soaking preformed NP-GLIDE coatings in a SO or SO mixture because of practicality considerations. Rather, we prepared PU-x%-y% or SINP-GLIDE coatings in the same way as the NP-GLIDE coatings except that the coating solution also contained a SO or SO mixture. Here y% denotes the weight fraction of SO with respect to GS and SO. The thickness of a coating was regulated by controlling the mass of the nonvolatile materials, including P1, P1-g-PDMS, HDIT, and SO, cast onto a glass plate. In our calculation of the thickness of a cross-linked coating from its mass, we used the density of 1.20 g/cm3 for pure PU since the PDMS content in the coatings never exceeded 12.0 wt %. To ensure the superior compatibility between P1 and P1-gPDMS and thus the uniform distribution of PDMS as nanopools in the cured coating, dimethyl carbonate (DMC) and DMF, which were selectively poor solvents for PDMS, were added into a coating solution in butanone prior to the casting of the solution (Supporting Information). The addition of DMC and DMF caused P1-g-PDMS to form micelles with coronas made of the P1 backbone and cores composed of PDMS and SO. We also determined that the addition of the slow-evaporating DMF facilitated the formation of a uniform and smooth coating. NP-GLIDE Coatings. The PU-x% or NP-GLIDE coatings prepared following the protocol described above were highly transparent. Figure S1 in the Supporting Information provides a plot of variations in the transmittance, at 500 nm, of 30 μm PUx% coatings as a function of x%. The transmittance values were always above 99% and barely changed from x = 3.0% to x = 12.0%. These coatings had higher transmittance values than our previous coatings2 because they were free of ungrafted PDMS chains, which could undergo macrophase separation from the coating matrix. As reported previously,2 the optical clarity arose because PDMS segregated into nanopools that were dispersed in the PU coating matrix and on the coating surface. Panels a−d of Figure 1 show AFM topography images of cross sections of PUx% coatings with x% = 3.0%, 6.0%, 9.0%, and 12.0%. Nanowells were visible in all of the four images. As x% increased, the density of the wells increased. Thus, the wells should be the locations of the PDMS nanopools. That appearance of PDMS domains as wells or pools was consistent with the finding of another group that used AFM and transmission electron microscopy in combination to study the morphologies of a PDMS-bearing block copolymer.39 The wells appeared probably because some of the PDMS chains under air escaped from the pools to cover the surfaces of the PU microtomed D

DOI: 10.1021/acsami.7b00126 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces in the measured T% values due to the nonuniform distribution of the macrophase-separated SO droplets in the coating. Figure 2c reveals an interesting trend that T% decreased more slowly with y% for coatings that were formed by evaporating most of the solvent at 45 ± 5 °C rather than at room temperature (22 ± 1 °C). This result might be due to the increased uptake capacity of SO by the GS nanopools at an elevated temperature. Figure 2d shows T% variation data for coatings at 45 ± 5 °C and containing a mixture of SO-1, SO-2, and SO-3 at the mass ratios of 2.00/2.00/1.00. In comparison with the data shown in Figure 2c, one sees that clearer coatings were obtained after SO-3 was replaced by the SO mixture. Additionally, the coatings containing the SO mixture and cast at 45 ± 5 °C were more transparent than those of the SO-1-containing coatings cast at room temperature. Thus, coatings for further studies were all cast at 45 ± 5 °C and contained this SO mixture. Morphologies of SINP-GLIDE Coatings. To investigate how SO impregnation affected the segregation of PDMS, we microtomed PU-6.0%-16.0% and PU-6.0%-32% samples and examined the cross-sectional surfaces via AFM. Figures 3a and 3b show the resultant AFM topography images. A comparison of Figures 1b, 3a, and 3b suggests that the nanopool density increased as y% increased at the fixed x% value of 6.0%. This density increase seemed to be accompanied by a reduction in the average pore sizes. While we should not overinterpret the latter observation because of the possibly different contributions made by the AFM tip to the measured pore sizes, the trends, if true, suggested that the incorporation of SO did not increase the size of the GS nanopools but instead yielded more nanopools, which helped solvating more SO without overstretching the GS chains. Antismudge Performance of SINP-GLIDE Coatings. Panels a and b in Figure 4 present snapshots taken out of video files that documented what happened after dyed hexadecane or water droplets were applied onto slanted glass slides that were not coated or coated by PU-6.0%-16.0%. While the hexadecane droplet spread and the water droplet got stuck on the untreated glass slide, both liquids cleanly glided down the coated slide without leaving a visible trace behind. In panels c and d, we note that the ink trace left by a Sharpie MAGNUM permanent marker and the spray from a ColorMaster KRYLON Banner Red paint contracted on our coating. Thus, the SINP-GLIDE coating, analogous to the NP-GLIDE coatings, possessed superior antismudge properties. Contact and Sliding Angles of Different Oils. We further compared the NP-GLIDE and SINP-GLIDE coatings by measuring the static contact angles (CAs) and sliding angles (SAs) of 5 μL oil droplets on the PU-6.0% and PU-6.0%-16.0% coatings for a wide range of test liquids that solubilized PDMS (group I) or did not solubilize PDMS (group II). In addition, the CAs on the unmodified base PU coating without P1-gPDMS were also measured. No SAs were reported for the PU base coating because the oils did not slide cleanly but spread on it similar to the situation depicted in Figure 4a. To minimize the effect of surface SO loss from previous tests on the measured CA and SA values, measurements were performed only on pristine locations of a coating sample. To generate a set of data, we measured CAs and SAs at 3−5 locations on a given sample and on three samples, and the data gathered were then used to calculate the averages and deviations (Table 1). A close examination of the data in Table 1 revealed various trends. First, the CAs of all oils on the base PU coating were

Figure 4. Snapshots cut from videos that recorded on a coated glass slide (a) the sliding of a dyed hexadecane droplet, (b) the sliding of a dyed water droplet, (c) the shrinking of a marker trace, and (d) the contraction of a sprayed paint. In each case, an uncoated slide was placed next to a coated slide for comparison.

lower than those on the PU-6.0% and PU-6.0%-16.0% coatings. Second, the CA and SA values declined as the surface tension γ of the test liquid decreased. For test liquids of group I, their SAs were always lower than 4° on the PU-6.0% and PU-6.0%-16.0% coatings. Third, a group II test liquid always had a higher CA and SA than its counterpart in group I with a similar γ. For example, methanol and decane have comparable γ values of ∼23 mN/m. The CA and SA for methanol were 44 ± 3° and 9 ± 2°, which were higher than the corresponding values of 15 ± 1° and 2 ± 1° for decane on the PU-6.0% coating. Fourth, while the SAs were essentially identical, within experimental error, for a given test liquid on the PU-6.0% and PU-6.0%16.0% coatings, the CA of a group I liquid was lower on the PU-6.0%-16.0% coating than on the PU-6.0% coating. Fifth, a critical γ* existed for the group I liquids. Below γ* ≈ 22 mN/ m, test liquids spread rather than cleanly glided on either the PU-6.0% or the PU-6.0%−16.0% coating. The first two trends can be explained by Young’s equation: γ − γLC cos θ = C γ (1) where γC is the surface tension of the coating and γLC is the interfacial tension between the test liquid and the coating. While the γC value for the unmodified PU coating should be ∼41 mN/m,41 the PDMS-modified coatings should have a γC value intermediate between 41 and 19.0 mN/m, the surface E

DOI: 10.1021/acsami.7b00126 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Table 1. Contact and Sliding Angles of 5 μL Droplets of Various Organic Liquids on PU, PU-6.0%, and PU-6.0%-16.0% Coatings PU-0% liquid

miscib with PDMS

γ (mN/m)

diiodomethane hexadecane THF dodecane decane octaneb SO-3 DMF methanolb ethanolb perfluorooctaneb

√ √ √ √ √ √ √ × × × ×

50.8 27.5 26.4 25.4 23.8 21.6 19.0 37.1 22.7 22.1 14.0

a

CA (deg) 41 8 7 4 3 2 2 32 4 4