U

WLa

b

a

ARRAA

KAHU

1

fltnfiiaw([Tfisc

tirTtf

0d

Progress in Organic Coatings 74 (2012) 794– 800

Contents lists available at SciVerse ScienceDirect

Progress in Organic Coatings

j ourna l ho me p ag e: www.elsev ier .com/ locate /porgcoat

V-resistant amorphous fluorinated coating for anodized titanium surfaces

alter Navarrinia,∗ , Maria Vittoria Diamantia , Maurizio Sansoteraa , Federico Persicoa , Wu Menghuaa ,uca Magagnina, Stefano Radiceb

Dipartimento di Chimica, Materiali e Ingegneria Chimica “Giulio Natta”, Politecnico di Milano, via Mancinelli 7, 20131 Milano, ItalySolvay Solexis, Research & Technology, Viale Lombardia 20, 20021 Bollate (MI), Italy

r t i c l e i n f o

rticle history:eceived 1 June 2011eceived in revised form 30 August 2011ccepted 21 September 2011vailable online 6 December 2011

a b s t r a c t

In this work, a high molecular weight copolymer of tetrafluoroethylene and perfluoro-4-trifluoromethoxy-1,3-dioxole (HYFLON® AD60) and two perfluoropolyethers (PFPEs) containingammonium phosphate or triethoxysilane functionalities, FLUOROLINK® F10 and FLUOROLINK® S10respectively, have been evaluated as protective coatings that can be easily applied on anodized tita-nium surfaces. Water and n-dodecane contact angle measurements have been recorded by using the

eywords:nodized titaniumYFLON® AD60V- stable fluorinated coating

sessile drop method for bare and coated surfaces in order to determine the hydrophobic and oleophobicproperties of the coatings. The UV-stability of coatings have been studied by Fourier transform infraredspectroscopy (FT-IR) analyses and by observing the variation of water contact angles on coated substratesbefore and after UV irradiation at regular time intervals. The thickness of the fluorinated films has beenmeasured by ellipsometry and by weight evaluation. Preliminary tests of the adhesion between films andsubstrate have been conducted.

. Introduction

In the past decade, titanium has received increased attentionrom scientists, engineers and architects, because of its pecu-iar properties like high strength-to-weight ratio and immunityo many corrosive environments [1]. In addition, through a tita-ium anodizing process it is possible to produce superficial oxidelms strongly adherent to the metallic substrate and character-

zed by dazzling brightness, unique corrosion resistance, appealingesthetics and more importantly photocatalytic surface [2–4]henever the oxide films are in the appropriate crystal phase

anatase, being the other crystal phases still active, yet less efficient)5–8]. Thanks to this photocatalytic activity of anodized titanium,iO2 has contributed to a wide variety of applications in severalelds including photocatalysis [9], environmental cleanup [10–12],olar energy conversion [13–19], electrochromic devices [20], self-leaning [21] and antimicrobial coating [22].

The photocatalytic reactions mostly take place on the surface;hat is, upon the UV irradiation on anodized titanium, the photo-nduced charge carriers might migrate to the surface where theyeact with adsorbed water and oxygen to produce radical species.

hese radicals would attack adsorbed organic molecules leadingo their complete decomposition into CO2 and H2O [23]. There-ore hydrogenated polymeric matrix usually applied as protective

∗ Corresponding author.E-mail address: walter.navarrini@polimi.it (W. Navarrini).

300-9440/$ – see front matter © 2011 Elsevier B.V. All rights reserved.oi:10.1016/j.porgcoat.2011.09.023

© 2011 Elsevier B.V. All rights reserved.

coating will ultimately be degradated upon UV irradiation, form-ing intermediates which could soil the anodized titanium surface[3,24]. Thus, finding UV-resistant polymeric coatings for photoac-tive titania would be a key step towards the practical application ofphotoactive anodized titanium in the fields of functional coatings.

Perfluoropolymers are characterized by very high thermaland chemical stability, low surface energies and low wettability[25–27]. These properties are useful to improve both dirt resistanceand washability of the anodized surface, since the fluoropolymercoating lowers the adhesion of staining agents to the coated surface[27]. Moreover, amorphous perfluoropolymers are highly transpar-ent materials in a very wide electromagnetic wavelength regionand are characterized by a very low refractive index [25,29,30].Chemically inert and highly transparent amorphous perfluoropoly-mers with low refractive index are successfully used to producepolymeric waveguides [31], and UV transparent thin films [30];therefore highly fluorinated polymers can be the salient candidatefor the creation of transparent and chemically stable coating foroptical application on anodized titanium [31].

In this paper, three coatings have been applied and eval-uated as resistant protective coatings on anodized titanium:perfluoropolymer HYFLON® AD60, i.e. a high molecular weightcopolymer of tetrafluoroethylene (TFE) and 2,2,3-trifluoro-5-trifluoromethoxy-1,3-dioxole (TTD) and two perfluoropolyethers

(PFPEs) containing ammonium phosphate and triethoxysilanefunctionalities, FLUOROLINK® F10 and FLUOROLINK® S10 respec-tively. The UV-stability of coatings has been studied by FT-IRanalyses and by measurements of water CAs on coated substrates

Organic Coatings 74 (2012) 794– 800 795

btoac

2

2

aHSamHaa1n–flfnSacgc(Pf2E

2

2

(ic[a

2

ltlHFF(

aafi

2

dn

Table 1Water and n-dodecane contact angle and apparent surface energy of uncoated andfluoropolymer-coated samples.

Sample Contact angle Apparent surfaceenergy (mN/m)

Water n-C12H14

Plate 0 49.12◦ 5◦ 46.16Plate 1 119.53 ± 0.22◦ 64.84 ± 3.32◦ 24.77Plate 2 90.53 ± 3.69◦ 66.67 ± 4.56◦ 25.89

W. Navarrini et al. / Progress in

efore and after UV irradiation at regular time intervals. Ellipsome-ry and weight evaluations have been used to measure the thicknessf the fluorinated coatings. Water and n-dodecane contact angles,s well as apparent surface energy, have been determined on theoated anodized titanium.

. Experimental

.1. Material

Commercially pure titanium metal (ASTM grade 2) was useds anodization support and amorphous perfluoropolymers (a)YFLON® AD60, (b) FLUOROLINK® F10 and (c) FLUOROLINK®

10 were tested as protective coatings, their molecular structuresre shown in Fig. 1. These polymers are produced and com-ercialized by Solvay-Solexis S.A. under the above trademarks:YFLON® AD60 is a random copolymer of tetrafluoroethylene (TFE)nd 2,2,4-trifluoro-5-trifluoromethoxy-1,3-dioxole (TTD) with anverage molecular weight (AMW) of 500,000 g/mol, Tg equal to30 ◦C; FLUOROLINK® F10 is a ˛–ω-ammonium phosphate termi-ated PFPE with an AMW of 2200 g/mol, the perfluoroether unitsOCF2– and –OCF2CF2– are randomly distributed along the per-uorinated polymeric backbone (RF) and the ending phosphate

unctionalities have a polar interaction with the anodized tita-ium surface increasing the adhesion of the coating; FLUOROLINK®

10 is a ˛–ω-terminated triethoxysilane PFPE derivative withn AMW of 1800 g/mol, the triethoxysilane functionalities at thehain-ending positions have the purpose to react with the –OHroups present on the anodized titanium surface, thus developinghemical bonds between coating and substrate. GALDEN® HT110Solvay Solexis) was used as solvent for HYFLON® AD60. It is aFPE-based solvent with a boiling point of 110 ◦C and formula asollow: CF3O(CF2CF(CF3)O)p(CF2O)nCF3 (p + n = 2–3; p/n = 20–50).-Propanol (99.8%, Sigma–Aldrich) and acetic acid (99.9%, Carlorba Reagents) were also used in the formulations of the coatings.

.2. Methods

.2.1. Titanium anodizationAt room temperature, acetone pre-degreased titanium plates

sizes: 15 mm width, 270 mm height, 1.5 mm thick) were anodizedn 0.5 M concentrated sulphuric acid by imposing a 40 mA/cm2

urrent and a 100 V cell voltage; as reported in previous work32], these anodizing parameters adopted would generate a violetnatase-enriched TiO2 film.

.2.2. Coating of anodized titanium platesBefore coating, three anodized plates were cleaned with distil-

ated water, acetone and a fluorinated solvent (i.e. CF3OCFClCF2Cl),hen the surfaces were separately brushed over with the fol-owing fluoropolymers solutions: Plate 1, 1% weight concentratedYFLON® AD60 in GALDEN® HT110; Plate 2, 1 wt% FLUOROLINK®

10 in isopropyl alcohol (20 wt%) and water (79 wt%); Plate 3, 1 wt%LUOROLINK® S10 in isopropyl alcohol (94 wt%), deionized water4 wt%) and acetic acid (1 wt%).

After coating, the plates have been dried with warm compressedir then underwent the following thermal treatment: holding inmbient atmosphere at 25 ◦C for 15 min, heating in oven to 100 ◦Cor 30 min, heating to 150 ◦C for 15 min and cooling back to 25 ◦Cn ambient atmosphere.

.2.3. EllipsometryEllipsometry has been used to evaluate the thickness of the

eposited films. Ellipsometric analyses were done with a Gaert-er ellipsometer L116S, equipped with a HeNe laser (� = 632.8 nm),

Plate 3 121.78 ± 1.32◦ 65.33 ± 4.27◦ 24.80

with a Glan Thompson calcite polarizer and a StokeMeter four-detector-photopolarimeter as analyzer. The size of the light beamwas about 15 �m and the incidence angle of the light was 70◦.The optical constants of the surface layers were calculated througha Maxwell Garnett Effective Medium Approximation (MG-EMA)[32–35]. The surface oxide layer before fluoropolymer coating hasbeen assumed as composed by a mixture of Ti + TiO2 with thicknessin the range 140–230 nm, with a complete absorption and a reflec-tion of the 632.8 nm incident light due to its extinction coefficientof 1.97. Optical constants were introduced into Fresnel equations tocalculate the amplitude ratio, � , the phase difference between lightwaves known as p- and s-polarized light waves, �, and the thick-ness of the fluoropolymer layer, searching the best fitting betweenexperimental and calculated values with a trial-and-error proce-dure [33,34].

2.2.4. UV stability test by FT-IR spectroscopy and CAsmeasurements

The coated anodized titanium plates have been placed in a UVRayonet reactor equipped with 8 mercury lamps emitting UV light(� = 300 nm) for 10 h. The coated samples have been character-ized by FT-IR spectroscopy before and after the UV irradiation atdefined time intervals. The photodegradation of polymeric coatingswas determined by monitoring the decrease in absorption inten-sity from FT-IR spectra (1024 scans) which were measured usingThermo Nicolet 380 equipped with Smart SAGA® (Smart AperturedGrazing Angle), spectra acquired under 2 cm−1 resolution. Irradia-tion and FT-IR acquisition were performed at room temperature.

The contact angles (CAs) measurements on coated\uncoatedsamples have also been recorded before and after UV light exposureas the same timing as above mentioned IR absorption monitoring.At room temperature, static CAs of water and n-dodecane weremeasured by sessile drop method in order to determine the appar-ent surface energy, hydrophobic and oleophobic properties of theperfluoropolymeric coatings. Therefore the adhesion of stainingagents on the coated surface as well as the stability of the fluo-rinated coating during the photocatalytic reaction under the UVlight irradiation have been qualitatively evaluated.

CAs measurements were performed with a “Dataphysics OCA”contact angle instrument and elaborated with “SCA20” software.The experimental average values and related standard variationsfor the treated samples (Plates 1–3) and for the simple anodizedtitanium sample (Plate 0) as internal reference were reported inSection 3.1, Table 1. The range of CAs was obtained from 4 individualmeasurements, each on a new spot.

2.2.5. Mechanical resistance testAfter anodization and application of the organic coating, the

treated titanium substrate was rubbed a counted number of times

by means of blotting paper with a contact surface of 2 cm2, under apressure of 63 kPa. The decrease in CAs as a function of the numberof rubbing cycles was measured.

796 W. Navarrini et al. / Progress in Organic Coatings 74 (2012) 794– 800

he coa

3

3

swvmtTgh1avwpsto

Twwpoiaaco

Fig. 1. Molecular structure of the amorphous fluorinated polymers used for t

. Results and discussion

.1. Contact angle (CA) measurement

As mentioned above, perfluoropolymeric coatings confer to theurfaces low surface energy and low wettability; materials treatedith these products become hydrophobic and assume water CA

alues higher than 90◦. In Table 1 the obtained CA data are sum-arized. The CAs of water (about 49◦) on the uncoated anodized

itanium (Plate 0) revealed the intrinsic hydrophilic property of theiO2 surface; in contrast, FLUOROLINK® S10 and HYFLON® AD60ive to the treated surface (Plate 3 and Plate 1) an extremely highydrophobicity, with average values of water CAs, respectively, of22◦ and 120◦. Samples treated with FLUOROLINK® F10 (Plate 2) arelso characterized by a hydrophobic surface having an average CAalue of 91◦, although hydrophobicity is less pronounced compareith the other two coatings due to its ionic, therefore hydrophilic,hosphate moieties. Shortly, CAs of water on the brush-coated TiO2urfaces are all greater than 90◦, offering a direct indication thathe hydrophobic fluorinated polymers were properly coated on thexide substrates.

n-Dodecane CAs analyses were also performed (see Table 1).he obtained values confirm the ability of coatings in lowering theettability of the treated surface also when the surface interactsith low paraffin compounds, thus giving the TiO2 surface oleo-hobic properties. n-Dodecane CA values are about 65◦ in the casef the coated samples, whereas on the uncoated TiO2 film the values lower than 5◦. As it can be seen by the CAs evaluated in water

nd in n-dodecane, the treated samples are contemporarily char-cterized by hydrophobic and oleophobic properties. With the dataollected, it was possible to evaluate the apparent surface energyf the samples which was noticeably reduced by the presence of

tings: (a) HYFLON® AD60, (b) FLUOROLINK® F10 and (c) FLUOROLINK® S10.

the organic coating with respect to the uncoated anodized sub-strate.

3.2. Coating thickness

The thickness of the fluoropolymer layer was estimated throughellipsometric measurements and compared to ponderal analysis.Data obtained through ellipsometry characterization of the sam-ples are shown in Table 2 and compared with the ponderal analysisresult which reported in Table 3. Calculated ellipsometric opticalconstants, � cal and �cal, are consistent with the experimental ones,� exp and �exp. The thickness values of the fluoropolymer layerscalculated through the ellipsometric model are slightly higher thanthose from the ponderal analysis. More in detail, coated titaniumanodized samples were weighed before and after the polymersdeposition. As the geometry of the plates is known, it is possi-ble to calculate an average coating thickness. Data collected arereported in Table 3 and appear to be comparable to the thick-nesses obtained by ellipsometric analyses. Therefore ellipsometryand ponderal analyses estimated by weight differences also showedthat coating thicknesses are quite similar in all cases, thus the coat-ing performances obtained are directly comparable.

3.3. Mechanical resistance

Due to the low surface energy, which improves the washa-bility of coated surfaces, perfluoropolymers often present lack ofadhesion to the substrate. Therefore, some preliminary tests were

made in order to estimate coating resistance to a mild mechani-cal stress drawn out in time by comparing the CAs change aftercoated anodized titanium was rubbed a counted number of timeswith blotting paper. The results, displayed in Fig. 2 show that there

W. Navarrini et al. / Progress in Organic Coatings 74 (2012) 794– 800 797

Table 2Ellipsometry data and coating thicknesses.

Sample Fluropolymer-coating � sp �sp � calc �calc Thickness (nm)

Plate 1 HYFLON® AD60 29.46 80Plate 2 FLUOROLINK® F10 33.38 74Plate 3 FLUOROLINK® S10 23.35 96

Fig. 2. Results of preliminary mechanical resistance test (Plate 1: anodized titaniumcF

iF1b(ttesehtaptFyttsagt–inbmtT

Nations Environment Program (UNEP) has announced that produc-

TP

oated with HYFLON® AD60, Plate 2: anodized titanium coated with FLUOROLINK®

10, and Plate 3: anodized titanium coated with FLUOROLINK® S10.

s a reduction of CAs for all the samples after 300 rubbing cycles.or HYFLON® AD60 and FLUOROLINK® S10-coated samples (Plate

and Plate 3), the CAs drop slightly during the first 30 cycles rub-ing, then CAs slowly decreased and tend to reach a plateau valuestill above 100◦) after 150 rubbing cycles. The interaction betweenhis polymeric matrix and the titanium oxide substrate is supposedo be low due to the perfluorinated nature of the coating, how-ver, we observed a fairly good adhesion of HYFLON® AD60 on theubstrate in this preliminary mechanical test. This result can bexplained by supposing that the observed good adhesion of thisigh molecular-weight and high Tg fluoropolymer is ascribed tohe polymeric cohesive interaction. These peculiar cohesive inter-ction properties characterize the HYFLON® AD60 coating as aolymer with good mechanical resistance and able to withstandhe stress imposed during the test at ambient temperature. AlsoLUOROLINK® S10 exhibits a good adhesion because of the hydrol-sis/condensation reactions among polymer terminal groups andhe –OH groups present on the substrate, allowing the forma-ion of covalent bonds between silane functional groups and theubstrate surface permitting both polymeric cohesive and surfacedhesive interactions. We expected FLUOROLINK® F10 to have aood adhesion on the surface of anodized titanium as well, thankso the adhesive interaction between the phosphate group and theOH groups present on the substrate surface, however, surpris-

ngly FLUOROLINK® F10 was constantly removed by increasing theumber of rubs as shown in Fig. 2. This behavior can be explainedy the fact that FLUOROLINK® F10 contains very low Tg and low

olecular-weight perfluoropolyether moieties, therefore leading

o a very low polymeric cohesive interaction between interfaces.he results of the preliminary mechanical resistance test show that

able 3onderal evaluation of the coating thicknesses and comparison with ellipsometry data.

Sample Coating

m (mg) d (mg/mm3) V

Plate 1 0.8 1.68 0Plate 2 0.9 1.73 0Plate 3 0.7 1.51 0

.32 30.07 82.35 200

.64 32.35 74.22 200

.00 23.51 95.81 170

intrinsic cohesive interaction of HYFLON® AD60 ensures improvedadhesion and higher mechanical proprieties with respect to theother tested coatings.

3.4. UV stability

By monitoring the variation of the absorption bands in FT-IRspectra, obtained from the samples coated with FLUOROLINK® F10and FLUOROLINK® S10 (Figs. 3 and 4), we noticed that the IR bandsin the spectra region from 1250 cm−1 to 1000 cm−1 significantlydecreased with increasing UV exposure time; These bands are inthe characteristic spectral region where normal modes involv-ing the stretching of C–F, C–C and C–O–C bonds are generallyobserved. So these bands are characteristic of the polymer chainbackbone [30,36]. These IR bands absorption decrement can beexplained by the decomposition of partially hydrogenate backboneof FLUOROLINK® F10 and FLUOROLINK® S10 coatings, that meansphotocatalytic oxidation of hydrogenated moieties was induced bythe photoactive anodized titanium in air. In contrast, the corre-sponding IR bands in the same spectra region of the sample coatedby HYFLON® AD60 which was exposed under the same frequencyUV irradiation showed no significant change with the UV exposingtime as shown in Fig. 5. These experimental evidences confirmedthe photocatalytic stability of HYFLON® AD60 fluoropolymer inpresence of activated TiO2 under the UV irradiation and thesubstantial instability of FLUOROLINK® F10 and FLUOROLINK®

S10. As already observed in the photooxidation of PFPE deriva-tives, the photo induced decomposition of FLUOROLINK® F10 andFLUOROLINK® S10 coatings was supposed to undergo a stepwiseoxidation, through the formation of fluorinated alcohols, aldehy-des and acids [37]. Surprisingly, even if these intermediates are verysensitive to infrared analysis, and none of them was observed by FT-IR during the UV stability tests; in contrast, only a linear decreaseof the infrared absorption which is attributed to the degradation ofthe fluorinated backbone chain was observed.

A possible explanation for these finding is that a rapid oxidationoccurred, rather than a stepwise oxidation, leading to the forma-tion of perfluorinated acids subsequently decomposed into carbondioxide and volatile low-molecular hydrofluorocarbon [38]. Theseunexpected findings greatly encouraged a further research aimed atclarifying the degradation mechanism and the suitable conditionsof the intermediate (perfluorinated acid), because perfluorinatedacids are known as bio-accumulating pollutant chemicals and arevery difficult to be degraded [39]. In fact, they can remain in thebiosphere for many years [40]. For this reason, in 2009 the United

tion and applications of perfluorinated acid should be curtailed.Therefore, a straightforward methodology to decompose thesepotentially dangerous chemicals into the less-harmful compounds,

Thickness

(mm3) Ponderal (nm) Ellipsometry (nm)

.476 176.4 200

.52 192.7 200

.464 171.7 170

798 W. Navarrini et al. / Progress in Organic Coatings 74 (2012) 794– 800

Fig. 3. FT-IR spectra of FLUOROLINK® S10 coated anodized titanium during UV exposure.

Fig. 4. FT-IR spectra of FLUOROLINK® F10 coated anodized titanium during UV exposure.

W. Navarrini et al. / Progress in Organic Coatings 74 (2012) 794– 800 799

ated

lce

mtFfleodi

Fl

Fig. 5. FT-IR spectra of Hyflon® AD60 co

ike carbon dioxide and hydrofluorocarbons in presence of photo-atalytic TiO2, is actually quite interesting from both a practical andnvironmental point of view.

These findings have also been corroborated by the CA measure-ents which were acquired on coated samples before and after

he exposure to UV irradiation at the same regular intervals (seeig. 6). This allowed a qualitative evaluation of the stability of theuorinated coating under UV light irradiation. As shown in Fig. 6,vident decrease of the CAs of water from the initial value measured

n Plate 2 and Plate 3 after exposure to UV irradiation, indicates theecomposition and partial mineralization of the fluorinated coat-

ngs as well as photoinduced activity of anodized titanium TiO2.

ig. 6. Water CA values measured on anodized titanium samples coated with FLUOROLIight.

anodized titanium during UV exposure.

Besides that HYFLON® AD60 coating showed a good resistanceto the photocatalytic oxidation on anodized titanium. This high Tg

amorphous perfluoropolymer is also characterized by possessinga high (33%) free volume fraction. That is because HYFLON® AD60high chain stiffness prevents the chain segments to pack closely,therefore nano-sized pores with diameter of around 10 A [41]are formed in the polymer. As consequence of the presence ofnanovoides, this polymer exhibits a good permeability to gases(particularly elementary gases such as H2, O2 and N2) and vapors;

its gas permeability is two orders of magnitude higher than thegas permeability of low free-volume-fraction glassy polymers,for instance polycarbonate [26]. They also show good selective

NK® F10, FLUOROLINK® S10 and HYFLON® AD60 at different exposure time to UV

8 Organ

pmsc

iHcHt

O

O

H

biFc

4

taboadhh

R

[

[

[

[[[

[

[

[

[

[

[[

[

[

[

[

[

[

[

[

[[

[

[

[

[

[

[

[

[40] N. Yamashita, et al., Perfluorinated acids as novel chemical tracers of global

00 W. Navarrini et al. / Progress in

ermeability to different organic vapors, such as small organicolecules, like methane, ethane or propane, while they show a

lightly lower permeability to polar gases and aromatic solventsompared to the above small organic molecules [27–29].

Because of this selective permeability, together with the chem-cal resistance and UV transparency/stability, the species like O2 or

2O (gas) can permeate through the amorphous perfluoropolymeroating and finally reach the surface of the activated TiO2. There,2O, OH− and H+ can react with electron–hole couple according

he following known reactions:

2 + e− → O2−

H− + h+ → OH•

2O + h+ → H2O2 + UV → 2 OH•

Then these highly oxidative species (O2−, OH• and H2O2) diffuse

ack through the perfluorinated coating and oxidize the pollut-ng agents located on the surface of the amorphous fluoropolymer.urther experiments to verify the photoactivity of HYFLON® AD60oated anodized titanium are ongoing.

. Conclusions

Among the above three fluorinated polymers, we observedhat the high Tg amorphous transparent Hyflon® AD60 revealed

good adhesion to the anodized titanium substrate, it conferredoth hydrophobicity and oleophobicity to the substrate, more-ver it resulted to have a good UV stability to the photoactivenodized titanium. In addition, a straightforward methodology toecompose perfluorinated carboxylic acids into carbon dioxide andydrofluorocarbons in presence of photocatalytic TiO2 has beenypothesized, further studies in this direction are under progress.

eferences

[1] S.R. Seagle, Titanium and titanium alloys, in: J.I. Kroschwitz (Ed.), The KirkOth-mer Encyclopedia of Chemical Technology, 24, fourth ed., John Wiley and Sons,New York, 1997, pp. 186–224.

[2] T.L. Thompson, et al., Studies of the photoactivation of TiO2—new photochem-ical processes, Chem. Rev. 106 (10) (2006) 4428–4453.

[3] W. Navarrini et al., US Patent 5,646,223.[4] A. Russo, et al., Perfluoro-4-methyl-1,3-dioxole: a new monomer for high-Tg

amorphous fluoropolymers, J. Fluorine Chem. 125 (1) (2007) 73–78.[5] J.L. Delplancke, et al., Self-colour anodizing of titanium, Surf. Technol. 16 (2)

(1982) 153–162.[6] A. Mills, et al., An overview on semiconductor photocatalysis, J. Photochem.

Photobiol. A: Chem. 108 (1) (1997) 1–35.[7] M.E. Sibert, et al., Electrochemical oxidation of titanium surface, J. Electrochem.

Soc. 110 (1) (1963) 65–72.[8] A. Fusjishima, et al., Titanium dioxide photocatalysis, J. Photochem. Photobiol.

C: Photochem. Rev. 1 (1) (2000) 1–21.

[9] T. Carlsono, et al., Photooxidation of methanol using vanadium pentox-

ide/titanium dioxide and molybdenum trioxide/titanium dioxide surface oxidemonolayer catalysts, J. Phys. Chem. 90 (1986) 5896.

10] A. Fujishima, et al., TiO2 Photocatalysis in Fundamentals and Applications, BKC,Inc, Tokyo, Japan, 1997.

[

ic Coatings 74 (2012) 794– 800

11] M.R. Hoffmann, et al., Environmental applications of semiconductor photo-catalysis, Chem. Rev. 95 (1995) 69.

12] J. Joo, et al., Large-scale synthesis of TiO2 nanorods via nonhydrolytic sol–gelester elimination reaction and their application to photocatalytic inactivationof E. coli, J. Phys. Chem. B 109 (2005) 15297–15302.

13] M. Grätzel, Photoelectrochemical cells, Nature 414 (2001) 338.14] M. Grätzel, Photochemistry reviews, J. Photochem. Photobiol. C 4 (2003) 145.15] A. Hagfeldt, et al., Light-induced redox reactions in nanocrystalline systems,

Chem. Rev. 95 (1995) 49.16] A.C. Arango, et al., Efficient titanium oxide/conjugated polymer photovoltaics

for solar energy conversion, Adv. Mater. 12 (2000) 1689.17] K. Hara, et al., A coumarin-derivative dye sensitized nanocrystalline TiO2 solar

cell having a high solar-energy conversion efficiency up to 5.6%, Chem. Com-mun. (2001) 569.

18] M.E. Butler, et al., Principles of photoelectrochemical, solar energy conversion,J. Mater. Sci. 15 (1980) 19.

19] K. Zhu, et al., Enhanced charge-collection efficiencies and light scattering indye-sensitized solar cells using oriented TiO2 nanotube arrays, J. Nano Lett. 7(2007) 69–74.

20] A. Ghicov, et al., Photoresponse in the visible range from Cr doped TiO2 nano-tubes, Chem. Phys. Lett. 433 (2007) 323–326.

21] R. Wang, et al., Light-induced amphiphilic surfaces, Nature 388 (1997) 431.22] Y. Kikuchi, et al., Photocatalytic bactericidal effect of TiO2 thin films: dynamic

view of the active oxygen species responsible for the effect, J. Photochem.Photobiol. A 106 (1997) 51–56.

23] U. Diebold, The surface science of titanium dioxides, Surf. Sci. Rep. 48 (2003)53–229.

24] W. Navarrini, et al., Anti-fingerprint fluorinated coating for anodized titaniumavoiding color alteration, J. Coat. Technol. Res 8 (2) (2011) 153–160.

25] A. Mills, et al., Novel TiO2 CVD films for semiconductor photocatalysis, J. Pho-tochem. Photobiol. A: Chem. 151 (9) (2002) 171–179.

26] V.S. Arcella, et al., A study on a perfluoropolymer purification and its applicationto membrane formation, J. Membr. Sci. 163 (2) (1999) 203–209.

27] I. Pinnau, et al., Gas and vapor transport properties of amorphous perfluorinatedcopolymer membranes based on 2,2-bistrifluoromethyl-4,5-difluoro-1,3-dioxole/tetrafluoroethylene, J. Membr. Sci. 109 (1) (1996) 125–133.

28] C. Hu, et al., Relationship between polymer structure and gas transport prop-erties in a series of fluorine-containing aromatic polyamide membranes foroxygen enrichment, Desalination 144 (1–3) (2002) 103–108.

29] H.S. Coombe, et al., Polymer membrane air separation performance for portableoxygen enriched combustion applications, Energy Convers. Manage. 48 (5)(2006) 1499–1505.

30] A. Milani, et al., Spectroscopic studies and first-principles modelling of2,2,4-trifluoro-5 trifluoromethoxy-1,3-dioxole (TTD) and TTD-TFE copolymers(Hyflon® AD), Polymer 49 (7) (2008) 1812–1822.

31] A. Lafuma, et al., Superhydrophobic states, Nat. Mater. 2 (7) (2003) 457–460.32] M.V. Diamanti, et al., Effect of anodic oxidation parameters on the titanium

oxides formation, Corros. Sci. 49 (2007) 939.33] V. Sirtori, et al., Calculation model of rough gold optical constants, Surf. Sci. 554

(2–3) (2004) 119–124.34] C.F. Bohren, et al., Absorption and Scattering of Light by Small Particles, Wiley,

New York, 1998.35] D.E. Aspnes, et al., Handbook of Optical Constant of Solids, Academic Press,

Orlando, 1985.36] J.B. Wachtman, Characterization of Materials, Manning Publications CO., Green-

wich, 1992, p. 158.37] J. Scheirs, et al., Photooxidation of functionalized perfluorinated polyethers,

Polym. Degrad. Stab. 56 (2) (1997) 239–253.38] R. Dillert, et al., Light-induced degradation of perfluorocarboxylic acids in the

presence of titanium dioxide, Chemosphere 67 (4) (2007) 785–792.39] J.P. Giesy, et al., Global distribution of perfluorooctane sulfonate in wildlife,

Environ. Sci. Technol. 35 (2001) 1339–1342.

circulation of ocean water, Chemosphere 70 (7) (2008) 1247–1255.41] W.J. Davies, et al., Positron annihilation lifetime and Doppler broadening

studies of amorphous fluoropolymer-poly(2,2-bis trifluoro 4,5 difluoro 1,3-dioxole), Eur. Polym. J. 30 (1994) 1289.