School Work

054_CRST_2011

Description
Biomaterials Paper
Categories
Published
of 5
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  On ice-releasing properties of rough hydrophobic coatings S.A. Kulinich ⁎ , M. Farzaneh 1 CIGELE /INGIVRE, Department of Applied Sciences, Université du Québec à Chicoutimi, 555 University Boulevard, Saguenay, PQ, Canada G7H 2B1 a b s t r a c ta r t i c l e i n f o  Article history: Received 6 October 2009Accepted 4 January 2010 Keywords: HydrophobicityRoughnessIce adhesion strengthDurability In this work, ice repellency of rough hydrophobic coatings based on different materials and with differentsurface topographies is evaluated. The coatings were prepared either from a  󿬂 uoropolymer incorporatedwith nanoparticles or by etching aluminum alloy substrate followed by further hydrophobization of therough surface via an organosilane monolayer adsorbed from solution. This allowed comparing the ice-releasing performance of rough surfaces with high water contact angles ( ∼ 150 – 153°) and different dynamichydrophobicities and mechanical properties. Arti 󿬁 cially created glaze ice, similar to naturally occurring glaze,was accreted on the surfaces by spraying supercooled water microdroplets in a wind tunnel at subzerotemperature. The ice adhesion strength was evaluated by spinning the samples in a centrifuge at constantlyincreasing speeds until ice detachment occurred. The results showed that, after several icing – deicing cycles,the more robust surfaces prepared by etching the aluminum substrate maintained their ice-releasingproperties better, compared to their counterparts based on nanoparticle-incorporated  󿬂 uoropolymer. Theeffect of the dynamic hydrophobicity of the coatings was also examined, clearly demonstrating that thesurface with low dynamic hydrophobicity is not ice-repellent, although it demonstrates large values of watercontact angle.Crown Copyright © 2010 Published by Elsevier B.V. All rights reserved. 1. Introduction Adhesionoficeandwetsnowtooutdoorsurfacesiswellknowntocause serious problems to power transmission lines, aircraft, boats,etc. (Croutch and Hartley, 1992; Andersson et al., 1994; Saito et al.,1997; Petrenko and Peng, 2003; Kako et al., 2004). Even though thereis no known material that completely prevents ice/snow buildups onits surface (Croutch and Hartley, 1992; Andersson et al., 1994;Kulinich and Farzaneh, 2004b), some coatings are believed to providereduced adhesion (Saito et al., 1997; Kulinich and Farzaneh, 2004a,2005,2009a;MeniniandFarzaneh,2009).Thisisexpectedtoresultinlower ice and/or wet-snow accumulation on such coated surfaces.Therefore, the research on coatings capable of reducing wet-snow,frost,oriceaccumulationhasbeengoingonfordecades(Muraseetal.,1994; Saito et al., 1997; Somlo and Gupta, 2001; Kako et al., 2002,2004; Ayeres et al., 2007a, 2007b; Wang et al., 2007; Kulinich andFarzaneh, 2004a, 2005, 2009b; Cao et al., 2009). Good correlationbetween hydrophobicity of surfaces and their ice-repellent behaviorwas previously reported by several groups (Saito et al., 1997;Petrenko and Peng, 2003). Others (Landy and Freiberger, 1967), however, found no correlation between the ice adhesion data andcontact angle (CA) on plastic surfaces.Superhydrophobic surfaces (i.e. those exhibiting water CA  N 150°)were  󿬁 rst tested by Saito et al. (1997), and demonstrated promisinganti-icingperformance.Veryrecently,MeniniandFarzaneh(2009)andSarkar and Farzaneh (2009) showed reduced ice adhesion onhydrophobiccoatingswithlargeCAvalues,whileKulinichandFarzaneh(2009a,b) demonstrated the effect of superhydrophobic surface CAhysteresis (CAH) on ice adhesion strength on such surfaces. Tourkineet al. (2009) and Cao et al. (2009) reported delayed water freezing onrough superhydrophobic surfaces, which is believed to be favorable forreducediceaccumulation.However,ice-releasingperformanceofroughsuperhydrophobic surfaces over time has not been reported to date.In this study, glaze ice was prepared by spraying water micro-droplets at subzero temperature, i.e. under conditions very close tooutdoor ice accretion. Ice adhesion was tested on rough super-hydrophobic samples with different topographies and based ondifferentmaterials(withdifferentchemicalandmechanicalproperties).Theirice-repellentpropertieswereexaminedafterseveralicing/deicingcycles, thus evaluating the durability of the samples. 2. Experimental section Aluminum alloy (AA6061-T6) plates, 3.2×5.0 cm 2 in size, wereusedas substratesfor all samples.Priortocoating, theywere polishedwith emery paper and cleaned in organic solvents. Sample 1 was aplate of etched aluminum alloy coated with organosilane (octadecyl-trimethoxysilane, ODTMS, from Aldrich). A plate was  󿬁 rst etched in17% HCl for 5 min, after which it was sonicated in deionized water, Cold Regions Science and Technology 65 (2011) 60 – 64 ⁎  Corresponding author. E-mail address:  skulinic@uqac.ca (S.A. Kulinich). 1 Tel.: +1 418 545 5044; fax: +1 418 545 5032.0165-232X/$  –  see front matter. Crown Copyright © 2010 Published by Elsevier B.V. All rights reserved.doi:10.1016/j.coldregions.2010.01.001 Contents lists available at ScienceDirect Cold Regions Science and Technology  journal homepage: www.elsevier.com/locate/coldregions  rinsed and dried in air for 1 h. To enable its superhydrophobicproperties,itwasthenimmersedin1%(v/v)water/methanolsolutionof ODTMS (9 ml of water and 90 ml of methanol). After 15 min, thesamplewasremovedfromtheODTMSbath,rinsedinmethanol,blow-dried with nitrogen  󿬂 ow and  󿬁 nally heat-treated overnight at ∼ 80 °Cin air. Samples 2 and 3 were prepared by following and modifyingrecipes previously reported in (Hsiang et al., 2007; Kulinich andFarzaneh, 2009c). TiO 2  nanopowder (particle size  b 50 nm) fromAldrich (8.0 g) was mixed with 80 ml of deionized water. Thesuspension was sonicated for 30 min, after which 5.0 ml of Zonyl8740 (a per 󿬂 uoroalkyl methacrylic copolymer product from DuPont)were added. The  󿬁 nal suspension was stirred for 3 h before beingcoated on the substrates. Superhydrophobic sample 3 was preparedby spraying the TiO 2  suspension over the substrate surface uniformlyandlettingitdryat ∼ 50 °C.Superhydrophobicsample2waspreparedby spin-coating the suspension on the substrate. Using etching orspray/spin-coating allowed preparing samples with different surfacetopographies and wetting hysteresis, as discussed below. Uponcoating, the samples were heat-treated at 120 °C in air for 3 h toremove residual solvents. The samples are brie 󿬂 y described in Table 1.Both CA and CAH values were measured on a Krüss DSA100contact-angle goniometer following standard procedures. The mea-surements were made at 23±0.5 °C, with 5-µL water droplets.Surface topographies were analyzed with a WYKO NT1100 opticalpro 󿬁 ler (from Veeco), atomic force microscopy (AFM, Escope fromVeeco), and scanning electron microscopy (SEM, JSM 6330-F from JEOL). X-ray photoelectron spectroscopy (XPS) was performed with aQuantum-2000 instrument from ULVAC-PHI.The ice adhesion tests were conducted on Al beams with samplesspuninahome-madecentrifugeapparatus(seeFig.1)placedinacoldroom at  − 10 °C. The samples attached to the beams were iced in awind tunnel at a wind speed of 10 m/s, temperature  − 10 °C, waterfeed rate of 2.5 g/m 3 , and average droplet size of   ∼ 80 µm (evaluatedby applying the collargol slide impact method, see e.g. Kollár andFarzaneh, 2009), resulting in glaze ice layers up to ∼ 1 cm thick over asample area of  ∼ 3.2×3.0 cm 2 (Fig. 1a). This ice geometry was enoughto avoid cohesion failure during spinning and provide well reproduc-ible results duringdeicing.Ice mass andarea werecarefully evaluatedboth before and after deicing. A counter-weight at the opposite endwas used to balance the beam with samples (Fig. 1b). The arti 󿬁 ciallyicedsampleswerespuninthecentrifugeplacedinaclimaticchamberat  − 10 °C to determine the rotational speed at which ice detachedfrom the sample surface. At the time of detachment (detected withsensorsembeddedintothecentrifugewalls),theadhesionstrengthof ice is assumed to be equal to the centrifugal force,  F  = mr  ω  2 , where  m is the ice mass,  r   is the beam radius, and  ω   is the rotational speed inrad/s. The shear stress, correspondingly, was calculated as  τ  = F  /  A ,where  A  is the deiced area. Mirror-polished aluminum sample(AA6061-T6 polished with 1  μ  m alumina slurry as the  󿬁 nal step)was used as a reference. Three pieces were prepared for each samplein Table 1, and the results were calculated as the average of the three.Further details on this technique can be found in recent reports(Menini and Farzaneh, 2009; Kulinich and Farzaneh, 2009a,b). 3. Results and discussion Table 1 brie 󿬂 y describes the samples used in this study, and bothCA and CAH values are presented. All the samples are seen todemonstrate CA values of water droplets within the range of   ∼ 150 – 153°. These values are characteristic of superhydrophobic surfaces.XPS analysis of samples 2 and 3 (not shown here) yielded resultsconsistent with those of previous studies (Kulinich and Farzaneh,2009a,b,c), implying that these samples had very similar surfacechemistry, and the nanoparticles were well covered by a  󿬂 uoropoly-merlayer.Thehigh CAandlowCAHvalues observedonsample1alsoled to assume good surface coverage of the sample with ODTMSmolecules.Surfaceimagesandpro 󿬁 lesofthesamplesarepresentedinFigs.2,3,and4forsamples1,2,and3,respectively.Allthesamplesareseentoberough at micro/nano-scale, with root-mean-square roughness ( R q )measured by AFM to be 241, 212, and 181 nm for samples 1, 2, and 3,respectively. Hence, air entrapment into such surface structures wasexpected during wetting. However, as well seen from the surfacepro 󿬁 les, sample 3 had a somewhat lower roughness, its surfaceasperities (Figs. 4b and 5b) were shorter and had relatively  󿬂 at andshallow tops, whereas those of samples 1 and 2 (Figs. 2a,b and 3a,b)weretaller,withsharperappearance,andmoreproperlyspaced.ThisisalsoseeninFig.5,wherethesurfacetopographies(measuredbyopticalpro 󿬁 ler) of samples 2 and 3 are compared. This difference in surfacetopographywasshowninourpreviousstudies(KulinichandFarzaneh,2009a,c) to resultin differentwetting modeson samples 2 and 3.Morespeci 󿬁 cally, while the Cassie – Baxter wetting mode was expected forsample2(andalsosample1whichalsohadahighersurfaceroughness),a mixed (Cassie – Baxter and Wenzel) mode was likely for sample 3(Kulinich and Farzaneh, 2009a,c). As a result, the water – solid contactareaonsample3wasexpectedtobelargerthanthoseonsamples1and2, which is in agreement with the contrasting wetting hysteresisobserved onthesesamples(see Table 1). This is clearlyshownin Fig. 6, wherethewettingofroughhydrophobicsurfacesinpureCassie – Baxter(top) and pure Wenzel (bottom) regimes is schematically illustrated.As natural icing events occur under more dynamic conditionsthan those previously applied for testing ice adhesion on materials(Andersson et al., 1994; Saito et al., 1997; Somlo and Gupta, 2001;  Table 1 Preparation and properties of the samples analyzed.Sample # Description Preparation CA (°) CAH (°)1 Etched Al/ODTMS Dip-coating 153.1±2.8 5.7±2.02 TiO 2 – Zonyl Spin-coating 152.2±2.3 6.1±2.13 TiO 2 – Zonyl Spraying 149.7±3.4  ∼ 80 Fig. 1.  (a) Coated sample attached to aluminum beam and covered with arti 󿬁 cial glazeice. (b) Coated sample in centrifuge set-up that evaluates ice adhesion: (1) sample,(2) aluminum beam, (3) counter-weight.61 S.A. Kulinich, M. Farzaneh / Cold Regions Science and Technology 65 (2011) 60 – 64  Petrenko and Peng, 2003; Sarkar and Farzaneh, 2009), this impliesthat the dynamic hydrophobicity of surfaces may play some role. Thisisbelievedtobeevenmoreimportantforroughhydrophobicsurfaceswhere air can be entrapped underneath water (e.g. Fig. 6, top).Therefore, similar to our previous works (Meniniand Farzaneh, 2009;Kulinich and Farzaneh,2009a,b), the glaze ice used to evaluatethe ice Fig. 2.  SEM surface image (a) and AFM surface pro 󿬁 le (b) of HCl etched Al coated withODTMS (sample 1). Fig. 3.  SEM surface image (a) and AFM surface pro 󿬁 le (b) of sample 2 based on 󿬂 uoropolymer incorporated with TiO 2  nanopowder. Fig. 4.  SEM surface image (a) and AFM surface pro 󿬁 le (b) of sample 3 based on 󿬂 uoropolymer incorporated with TiO 2  nanopowder. Fig.5. Surfaceimagesofsamples2(a)and3(b)takenbyopticalpro 󿬁 ler.Thesampleswereprepared from TiO 2 – Zonyl suspension and demonstrate low (a) and high (b) values of wetting hysteresis.62  S.A. Kulinich, M. Farzaneh / Cold Regions Science and Technology 65 (2011) 60 – 64  repellency of the samples was prepared by spraying water micro-droplets in a wind tunnel maintained at subzero temperature.Fig. 7a presents shear stress of ice detachment values of thesamples as a function of the number of icing/deicing cycles, and theaverage value obtained on uncoated mirror-polished aluminum( ∼ 360 kPa) is also indicated. In agreement with our previous workon low-CAH superhydrophobic surfaces (Kulinich and Farzaneh,2009a,b), the initial values of shear stress of ice detachment onsamples 1 and 2 were  ∼ 3.5 – 4.4 times lower than on the polishedaluminum standard (being  ∼ 80 and 100 kPa, samples 1 and 2 inFig. 7a), which is consistent with the above mentioned Cassie – Baxterwettingregimeontheirsurfaces(seeFig.6,top)thateventuallyledtosmall ice – solid contact areas (Kulinich and Farzaneh, 2009a,b). Aftersix icing/deicing experiments, however, the samples demonstratedsomewhat different behavior. While ice adhesion strength on sample1 did not change signi 󿬁 cantly (squares in Fig. 7a), that on sample 2(circles)increasedbyafactorofabouttwo.Theinitialvalueonsample3 (triangles) was signi 󿬁 cantly larger than those on samples 1 and 2,and even slightly larger than that on the uncoated aluminum stan-dard. This agrees well with previous work (Kulinich and Farzaneh,2009a,b) and is believed to be associated with a larger ice – solidcontact area on this surface (which was inherited from a mixedCassie – BaxterandWenzelwettingmodeby wateronthis surface).Asthe number of icing/deicing cycles increased, however, the initialvalue remained essentially unchanged after as many as six icing/deicing events (Fig. 7a, sample 3). Hence, ice adhesion strengthobserved on the high-CAH sample 3, being initially high, did notchange much after six icing/deicing events.To explain these  󿬁 ndings, we performed AFM surface analyses of the samples after deicing experiments in order to follow changes (if any) in sample topographies. Root-mean-square roughness ( R q ) wasevaluated over several different locations on each sample, and theresults are presented in Fig. 7b as a function of icing/deicing events.Only a small initial decrease in surface roughness is observed forsample 3 (triangles in Fig. 7b), which can be associated with someminor increase of ice adhesion strength on this surface during thesubsequenticing/deicingcycles(Fig.7a).Similarly,asmalldecreaseinsurfaceroughnessappearstobeseenforsample1(squaresinFig.7b),and this corresponds well to the essentially unchanged ice adhesionstrengthonthissampleoversixicing/deicingcycles(Fig.7a,squares).Meanwhile, the change in surface roughness for sample 2 (circlesin Fig. 7b) is more remarkable (from  ∼ 210 to  ∼ 175 nm), and thiscorrelates well with the increase in ice adhesion strength on thissample, as observed in Fig. 7a (circles).The rough superhydrophobic surfaces in this study are thereforeshown to demonstrate different ice-releasing performance. The high-CAH surface (sample 3) demonstrated ice adhesion strength compa-rable or greater than that on polished uncoated aluminum. However,it did not appear to change over six icing/deicing events. This isbelieved to be related to the relatively  󿬂 atter surface topography of the sample (Figs. 4 and 5b), which was not signi 󿬁 cantly damagedduringicing/deicing.The sharper andmuchtallerasperitiesin sample2(Figs.3and5a)werebelievedtobegraduallydamagedduringicing/deicing. Their sharp tips could be broken and removed by ice duringdeicing. On the other hand, some damage could be caused by icing aswell, since water expanded on the solid surface on freezing (inducinga very signi 󿬁 cant interfacial stress). This is con 󿬁 rmed by the gradualdecrease in surface roughness observed for this sample in Fig. 7b.Sample 1 (also with sharp and relatively tall asperities) seemed,however, to be more resistant to damage during icing/deicing. Thiswas likely due to its more rigid asperities (built of Al  x O  y ) compared tothose in sample 2 (based on  󿬂 uoropolymer heavily loaded with TiO 2 nanoparticles).The anti-ice performance of rough hydrophobic surfaces is thusdemonstratedtobedependentonthenumberoficing/deicingevents.Sharp and tall surface asperities on such surfaces can be damagedor broken during icing/deicing, gradually leading to reduced ice-releasing performance. And therefore the mechanical properties of such materials must be taken into account in order to further developthem for anti-icing applications. 4. Conclusions The adhesion of glaze ice, which was prepared from supercooledwater droplets similar to ice accreted in nature, was measured onrough hydrophobic surfaces as a function of icing/deicing cycles. The Fig. 6.  Wetting of rough hydrophobic surfaces in the Cassie – Baxter (top) and Wenzel(bottom) regimes. Air is entrapped underneath water drop on the top surface. Fig. 7.  Shear stress of ice detachment (a) and root-mean-square roughness (b) asfunction of icing/deicing cycles. Dotted lines are only guide to the eye, and dashedhorizontal line in (a) at ∼ 360 kPa indicates average value obtained on mirror-polishedaluminum.63 S.A. Kulinich, M. Farzaneh / Cold Regions Science and Technology 65 (2011) 60 – 64
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks