Langmuir 2007, 23, 209-216
209
Water Makes It Hydrophobic: Contraphilic Wetting for Polyurethanes with Soft Blocks Having Semifluorinated and 5,5-Dimethylhydantoin Side Chains† Umit Makal, Nergis Uslu, and Kenneth J. Wynne* Department of Chemical and Life Science Engineering, Virginia Commonwealth UniVersity, Richmond, Virginia 23284 ReceiVed May 31, 2006 A series of polyurethanes with novel copolymer soft blocks display a new surface phenomenon, contraphilic wetting, in which the dry surface is hydrophilic and the wetted surface is hydrophobic. A precursor polymer was prepared with copolymer soft blocks containing semifluorinated (trifluoroethoxy, 3FOx, or pentafluoropropoxy, 5FOx) and bromomethyl functional pendant groups with 2:1, 1:1, and 1:2 semifluorinated/bromomethyl ratios. The hard block consists of isophorone diisocyanate (IPDI) and 1,4-butanediol (BD). 5,5-Dimethylhydantoin was introduced by the substitution of Br via reaction-on-polymer. The composition, structure, and percent of 5,5-dimethylhydantoin substitution for both the precursor and the 5,5-dimethylhydantoin-substituted polyurethanes were analyzed by 1H NMR. The difference between the advancing contact angle on the wetted surface and that on the dry surface (∆C) is highest (38°) for the polyurethane with the highest ratio of semifluorinated/hydantoin soft block side chains. A model is proposed according to which contraphilic wetting is driven enthalpically by hydrogen bonding. For the dry surface, hydrogen bonding of 5,5-dimethylhydantoin amide carbonyl groups to methylene hydrogens of semifluorinated groups disrupts the normal surface concentration of semifluorinated groups, whereas the geometric arrangement of hydantoin N-H results in availability for hydrogen bonding with water. Upon exposure to water, amide groups switch from hydrogen bonding to -CH2CF2CF3 to stronger hydrogen bonding with water. As a result, semifluorinated groups are “released”, and the surface becomes hydrophobic. Drying the coating (50 °C) reversibly restores hydrophilic character. Coatings stored at ambient temperature and humidity have ∆C values intermediate between dry and wet states.
Introduction The wetting behavior of polymer surfaces is of fundamental importance for diverse applications such as paints and coatings, microfluidics, cosmetics, and biomaterials. The outermost molecular layer controls the surface response to external stimuli such as the initial wetting response to water. Thus, changes on the micro- to nanoscale may be employed to control macroscale wetting behavior.1-5 For example, we have recently shown that polyurethanes having copolymer soft blocks with random and block sequences result in different nanoscale surface morphologies and different macroscale wetting behavior.6 Although time-invariant surface properties are often sought, responsive surfaces may be desired that switch behavior in response to external stimuli. Several groups have designed nanofilms that switch between hydrophilic and hydrophobic states.7-10 Whether Y-shaped molecules8 or copolymer brushes,10 the surface mimics the polar or nonpolar characteristics of the last solvent to impinge on the surface. This behavior follows that anticipated on the basis of solubility parameters. † Part of the Stimuli-Responsive Materials: Polymers, Colloids, and Multicomponent Systems special issue.
(1) Holmes-Farley, S. R.; Reamey, R. H.; Nuzzo, R.; McCarthy, T. J.; Whitesides, G. M. Langmuir 1987, 3, 799-815. (2) Sun, T. L.; Wang, G. J.; Feng, L.; Liu, B. Q.; Ma, Y. M.; Jiang, L.; Zhu, D. B. Angew. Chem., Int. Ed. 2004, 43, 357-360. (3) Russell, T. P. Science 2002, 297, 964-967. (4) Khongtong, S.; Ferguson, G. S. J. Am. Chem. Soc. 2002, 124, 7254-7255. (5) Carey, D. H.; Grunzinger, S. J.; Ferguson, G. S. Macromolecules 2000, 33, 8802-8812. (6) Fujiwara, T.; Wynne, K. J. Macromolecules 2004, 37, 8491-8494. (7) Julthongpiput, D.; Lin, Y. H.; Teng, J.; Zubarev, E. R.; Tsukruk, V. V. J. Am. Chem. Soc. 2003, 125, 15912-15921. (8) Luzinov, I.; Minko, S.; Tsukruk, V. V. Prog. Polym. Sci. 2004, 29, 635-698. (9) Julthongpiput, D.; Lin, Y. H.; Teng, J.; Zubarev, E. R.; Tsukruk, V. V. Langmuir 2003, 19, 7832-7836. (10) Granville, A. M.; Boyes, S. G.; Akgun, B.; Foster, M. D.; Brittain, W. J. Macromolecules 2004, 37, 2790-2796.
Poly(N-isopropylacrylamide) undergoes a phase transition with a lower critical solution temperature (LCST) close to 32°.11 The surface thermally switches between a hydrophilic state below the LCST and a hydrophobic state above the LCST.2 Some surfaces have been designed to change wetting behavior in response to an electrical potential.12 These nanofilms undergo conformational transitions between a hydrophilic and moderately hydrophobic state in response to an electrical stimulus.12 A counterintuitive case is increased hydrophobicity for oxidized 1,2-polybutadiene surfaces exposed to hot water. It was suggested that in lightly cross-linked systems rubber elasticity competes with the thermodynamic tendency of the surface to minimize surface free energy. As a consequence, the functional groups attached to the mobile segments stretch out of their random coil conformations. When the temperature is increased, the entropic loss translates into a restoring force and the chains recoil, pulling the hydrophilic functional groups away from the air-polymer interface and gradually increasing the advancing contact angle. On cycling, the surface eventually remains hydrophilic independent of temperature.4,5,13 We recently reported a novel responsive surface that reversibly switched between a hydrophilic state when dry and a hydrophobic state when wet.14 The term “contraphilic” was suggested for this new wetting behavior because the coatings behave in a way opposite to the anticipated amphiphilic response. The unusual surface properties were found during studies of wetting behavior for precursors to biocidal polymeric surface modifiers.15 The polyurethanes that display contraphilic behavior have copolymer soft (11) Bae, Y. H.; Okano, T.; Kim, S., W. J. Polym. Sci., Part B: Polym. Phys. 1990, 28, 923-936. (12) Lahann, J.; Mitragotri, S.; Tran, T.-N.; Kaido, H.; Sundaram, J.; Choi, I. S.; Hoffer, S.; Somorjai, G. A.; Langer, R. Science (Washington, D.C.) 2003, 299, 371-374. (13) Khongtong, S.; Ferguson, G. S. J. Am. Chem. Soc. 2001, 123, 35883594. (14) Makal, U.; Wynne, K. J. Langmuir 2005, 21, 3742-3745.
10.1021/la0615600 CCC: $37.00 © 2007 American Chemical Society Published on Web 08/08/2006
210 Langmuir, Vol. 23, No. 1, 2007
Makal et al.
Figure 1. 1H NMR spectrum of (A, CDCl3) IPDI-BD(40)/P(3FOx/BrOx-1:1) and (B, DMSO-d6) IPDI-BD(40)/P(3FOx/HyOx/BrOx-1.0: 0.65:0.35)
blocks containing semifluorinated and hydantoin side chains (Figure 1). To gain some understanding of contraphilic wetting, polyurethanes with copolymer soft blocks having different ratios of semifluorinated/5,5-dimethylhydantoin side chains have been prepared and evaluated by dynamic and static contact angle measurements. These studies have been augmented by compositional and (15) Makal, U.; Wood, L.; Ohman, D. E.; Wynne, K. J. Biomaterials 2006, 27, 1316-1326.
thermal characterization using 1H NMR, DSC, and GPC. In the course of these studies, it was found that an annealing step amplified the contraphilic behavior. Our results are reported below, together with a suggested mechanism for contraphilic wetting behavior. Experimental Section Materials. 3-(2,2,2-Trifluoroethoxymethyl)-3-methyloxetane (3FOx), 3-(2,2,3,3,3-pentafluoropropoxymethyl)-3-methyloxetane (5FOx), and 3-bromomethyl-3-methyloxetane (BrOx) were syn-
Contraphilic Wetting for Polyurethanes Scheme 1. 5,5-Dimethylhydantoin Substitution Reactiona
a (A) Precursor polyurethane; (B) 5,5-Dimethylhydantoinsubstituted polyurethane.
thesized following published procedures16 or were generously provided by Gencorp Aerojet (Sacramento, CA) or OMNOVA Solutions (Akron, OH). Monomers were distilled under vacuum before use: 3FOx and 5FOx at 100°C/5 mmHg and BrOx at 85°C/5 mmHg. Boron trifluoride diethyl etherate (BF3O(C2H5)2) was purchased from Aldrich and used as received. 5,5-Dimethylhydantoin (Aldrich, 97%), potassium carbonate (Acros Chemicals, ACS), and sodium thiosulfate (Aldrich, 99%) were used as received. Methylene chloride (anhydrous), tetrahydrofuran (THF, ACS), dimethylformamide (DMF, anhydrous), and methanol (ACS) were purchased from Aldrich and either used as received or dried and stored over 4 Å molecular sieves. Isophorone diisocyanate (IPDI, 98%), poly(tetramethylene oxide) (PTMO-2000), CDCl3, DMSO-d6 and dibutyltin dilaurate catalyst (T-12) were obtained from Aldrich. 1,4Butanediol (BD) was purchased from Acros Chemicals and used as received. Synthesis. The synthesis and characterization of the P(FOx/BrOx) telechelics and IPDI-BD/P(FOx/BrOx) polyurethanes have been reported.17 IPDI-BD/P(FOx/BrOx) polyurethanes were typically purified by reprecipitation. Polymer solutions in THF were added dropwise to (4:1) deionized water/methanol. The 1H NMR (CDCl3) for a representative precursor PU, IPDIBD/P(FOx/BrOx-1:1), is given in Figure 1A. The 1:1 FOx/BrOx composition is confirmed by the ratio of a (1.07 ppm, CH3, BrOx) to b (0.88 ppm, CH3, 3FOx). The FOx/BrOx ratio is more clearly delineated in the 1H NMR spectrum of the telechelics reported previously.17 Other peaks (ppm vs TMS) with assignments made according to the structure in Figure 1A were 0.8-0.9 ppm (m, n, CH3, IPDI); 1.7 ppm (h, -OCH2CH2CH2CH2O-); 3.1-3.3 ppm (f, CH2 methylenes for the three segments); 3.45 ppm (e, CH2-Br); 3.48 ppm (d, -OCH2CF3); 3.7-3.8 ppm (c, -CH2-O-CH2CF3); and 4.0-4.1 ppm (g, -OCH2CH2CH2CH2O-). Substitution of 5,5-dimethylhydantoin on the precursor polyurethanes was carried out in DMF in the presence of K2CO3 (Scheme 1). In a typical procedure, 5.32 g (38.50 mmol) of K2CO3 was dispersed in 13.56 mL of DMF with 0.98 g (7.65 mmol) of 5,5dimethylhydantoin in a 100 mL flask. The mixture was heated to 65°C under a nitrogen purge with a condenser. A solution containing 4.12 g (0.61 mmol of soft block) of precursor PU in 9.80 mL of DMF was added dropwise to the flask, and the temperature was increased and kept at 90-95°C. After 72 h, the reaction was quenched by decreasing the temperature to ambient. The product was precipitated into deionized water, filtered, and dried at 60 °C under vacuum. Evidence for the extent of 5,5-dimethylhydantoin substitution on the soft block of the polyurethanes was verified by the 1H NMR (16) Malik, A. A.; Archibald, T. G.; GenCorp: U.S. Patent 6,037,483, 2000. (17) Makal, U.; Uilk, J.; Kurt, P.; Cooke, R. S.; Wynne, K. J. Polymer 2005, 46, 2522-2530.
Langmuir, Vol. 23, No. 1, 2007 211 spectrum. As an example, the structure and corresponding 1H NMR spectrum (DMSO-d6) for IPDI-BD(40)/P(3FOx/HyOx/BrOx-1.0: 0.65:0.35) is shown in Figure 1B. The ratio of peak l (1.23 ppm (CH3)2, 5,5-dimethylhydantoin) to b (0.8-0.9 ppm, -CH3 (3FOx)) was used to determine the extent of hydantoin substitution. The result was checked by determining the ratio of peak l to peak d (3.46 ppm -OCH2CF3 (3FOx)). Other peaks (ppm vs TMS) with assignments according to the structure in Figure 1B were 0.7-0.8 ppm (j, CH3, 5,5-dimethylhydantoin-containing segment); 0.9-1.0 ppm (a, m, n, CH3, BrOx; CH3, IPDI); 1.55 ppm (h, OCH2CH2CH2CH2O-); 3.0-3.4 ppm (c, e, f, and k, which are methylenes for the three segments); 3.9-4.0 ppm (g, -OCH2CH2CH2CH2O-); and 8.2 ppm (p, CONH, 5,5-dimethylhydantoin). Characterization. Polyurethane (DMSO-d6) 1H NMR spectra were recorded using a Varian spectrometer (Inova 400 MHz) operating at 400 MHz. The FTIR spectra were obtained on a Nicolet 400 FTIR spectrometer using solution-cast films on KBr disks. Differential scanning calorimetry (DSC) was carried out on a TA-Q 1000 series instrument (TA Instruments). The DSC samples were directly deposited onto the DSC pan. Unless otherwise noted, measurements were carried out at a heating rate of 10 °C/min from -75 °C under an inert atmosphere. Indium metal was used for calibration. In addition to standard DSC, temperature-modulated DSC (MDSC) with a modulation amplitude of (0.5 °C, modulation period of 60 s, and heating rate of 3 °C/min was also performed. Molecular Weight Determination. Polyurethane molecular weights were measured using a Viscotek TriSEC triple detector GPC system (THF) with sample concentrations of 10-15 mg/mL and a flow rate of 1.00 mL/min. Universal calibration by polystyrene standards was used to determine the molecular weight (Mn, Mw) and polydispersity. Wetting Behavior. Dynamic contact angle (DCA) analysis based on the Wilhelmy plate method18 was carried out with a Cahn model 312 analyzer (Cerritos, CA). The surface tension quantification limit of the instrument is 0.1 dyn/cm. The probe liquid was ∼18 MΩ cm deionized water from a Milli-Q (Millipore) Nanopure system. The surface tension of the probe liquid was checked daily and was typically 72.6 ( 0.5 dyn/cm. Beakers used for DCA analysis were cleaned by soaking in an 2-propanol/potassium hydroxide base bath for at least 24 h, rinsed for 30 s with hot tap water, and then rinsed another 30 s with Nanopure water. In a typical determination, a coated slide was attached to the electrobalance via a clip, and a beaker of water was placed on the stage. The stage was automatically raised and lowered to allow water to impinge upon the slide. By analyzing the resulting force versus distance curves (fdc’s), advancing (θadv) and receding (θrec) contact angles were obtained. Unless otherwise noted, the stage speed was 100 µm/s, and the dwell time between advancing and receding test segments was 10 s. The static contact angles and image profiles were obtained by using a Rame´-Hart goniometer equipped with a camera. The contact angles were calculated using Drop Image software (version 1.4.11).
Results and Discussion The incorporation of 5,5-dimethylhydantoin (Scheme 1) was initially of interest for biocidal polymer surface modifiers (PSMs). Conversion of the amide functionality of the 5,5-dimethylhydantoin moiety to chloramide with dilute hypochlorite solution results in oxidative biocidal surfaces.19,20 We have verified the biocidal activity of polyurethane coatings containing 2 wt % 5,5-dimethylhydantoin PSMs against both gram-positive-type (S. aureus) and gram-negative-type (P. aeruginosa and E. coli) bacteria.15 During surface characterization of 5,5-dimethylhydantoin containing polyurethane intermediates, we observed unusual (18) Wilhelmy, L. Ann. Phys. Chem. Leipzig 1863, 119, 177. (19) Eknoian, M. W.; Worley, S. D. J. Bioact. Compat. Polym. 1998, 13, 303-314. (20) Worley, S. D.; Sun, G. Trends Polym. Sci. 1996, 4, 364-370.
212 Langmuir, Vol. 23, No. 1, 2007
Makal et al.
Table 1. Soft Block Compositions and Degree of 5,5-Dimethylhydantoin Substitution
polyurethanes were used as substrates for hydantoin introduction as described below. Hydantoin Containing Polyurethanes. 5,5-Dimethylhydantoin containing polyurethanes were obtained via a substitution reaction on IPDI-BD(40)/P(FOx/BrOx-m:n) polyurethanes (Scheme 1).21 A description of a typical procedure is given in the Experimental Section. The optimum reaction time was about 72 h (60-65% substitution). Longer reaction times resulted in only modest increases in substitution (e.g., 96 h, 66-70%). Furthermore, longer reaction times resulted in degradation as the reaction mixture turned dark brown. The slow rate and incomplete substitution of CH2-Br by hydantoin was likely due to steric hindrance at the neopentyl carbon site. The degree of hydantoin substitution was obtained by comparing the peak for methyl groups on the semifluorinated segment (0.86 ppm, DMSO-d6) to dimethyl peaks for 5,5-dimethylhydantoin (1.23 ppm, DMSOd6). The complete 1H NMR assignments for a representative precursor and 5,5-dimethylhydantoin containing polyurethane are given in Figure 1. The additional methyl and amide proton peaks on PU-2 at 1.23 and 8.21 ppm, respectively, are due to 5,5-dimethylhydantoin. Methylene peaks due to the HyOx segment are not well resolved because of overlap with methylene protons on other segments. The percent 5,5-dimethylhydantoin substitution is given in Table 1 along with soft block compositions. Two different series of polyurethanes were prepared having trifluoroethoxy, 3FOx, (PU-1, -2, and -3) or pentafluoropropoxy, 5FOx, (PU-4, -5, and -6) as the semifluorinated pendant group. The ratio of semifluorinated groups to hydantoin was controlled by using precursor polyurethanes with 2:1, 1:1, or 1:2 semifluorinated/bromomethyl ratios. The compositions of the resulting 5,5-dimethylhydantoinsubstituted polyurethanes are designated IPDI-BD(40)/P(3FOx/ HyOx/BrOx-x:y:z), where HyOx is the 5,5-dimethylhydantoincontaining segment. Molecular Weights. Soft block compositional designations and molecular weights of both precursor and 5,5-dimethylhydantoin-substituted polyurethanes are given in Table 2. GPC analysis gave Mw’s in the range of (30-47) × 103 g/mol for the precursors and (25-40) × 103 g/mol for the hydantoin-substituted polyurethanes. Thus, the molecular weights of precursor and hydantoin-substituted polyurethanes are about the same or slightly diminished. The soft block molecular weight is increased through the substitution of Br (80 g/mol) by 5,5-dimethylhydantoin (128 g/mol), but the expected increase in molecular weight is not detected by GPC analysis. Some slight chain cleavage during the substitution reaction may have compensated for the expected increase in molecular weight. Differential Scanning Calorimetry. For polyurethanes, two glass-transition temperatures (Tg’s) due to soft and hard block domains are usually detectable. Herein, thermal analysis was performed using both temperature-modulated (MDSC) and
(FOx)x(HyOx)y(BrOx)z soft segment composition
initial FOx/BrOx
% Br replaced
x
y
z
3FOx/BrOx
2:1 1:1 1:2
60 65 45
2.0 1.0 1.0
0.60 0.65 0.90
0.40 0.35 1.10
5FOx/BrOx
2:1 1:1 1:2
68 60 55
2.0 1.0 1.0
0.68 0.60 1.10
0.32 0.40 0.90
wetting behavior. The hydrophilic dry coatings (θadv, 83°) became hydrophobic (θadv, 110°) when wet. The term “contraphilic” was suggested for this wetting behavior because it was opposite to an expected amphiphilic surface response.14 Upon drying the coating at 50 °C under vacuum, the initial hydrophilic state was restored. In this report, the soft block compositional dependence and the effect of annealing on contraphilic wetting are investigated. For contraphilic wetting, our only successful preparative route so far involves a reaction-on-polymer displacement of Br by 5,5-dimethylhydantoin (Scheme 1). An oxetane monomer corresponding to the HyOx segment is readily prepared, but ringopening polymerization fails. Replacement of Br by 5,5dimethylhydantoin on telechelics is successful, but attempted incorporation of the substituted telechelics into polyurethanes resulted in gelation. Below, we briefly describe the characterization of precursor polyurethanes that have copolymer soft blocks composed of semifluorinated and bromomethyl side chains. This is followed by a discussion of the characterization and wetting behavior of polyurethanes with 60-70% of Br replaced by 5,5dimethylhydantoin. Precursor Polyurethanes. Copolymerization of 3FOx and BrOx was carried out by a modification of a published procedure.16,17 Cationic ring-opening polymerization of 3FOx and BrOx using BF3 etherate catalyst and 1,4-butanediol (BD) cocatalyst gave telechelics with compositions close to the feed ratio (Table 1). The telechelic (soft block) designation is P(FOx/ BrOx-m:n), where the segment ratios follow the 3FOx or 5FOx and BrOx repeat units and P designates the ring-opened structure in the telechelic. The presence of alcohol end groups (-CH2C(CH3)(R)CH2OH) is understood when the P(FOx/BrOxm:n) representation is used for the telechelics. Isophorone diisocyanate (IPDI), 1,4-butanediol (BD), and P(FOx/BrOx-m: n) gave IPDI-BD(40)/P(FOx/BrOx-m:n) polyurethanes in a conventional soft-block-first procedure.17 Precursor polyurethanes with varying hard block content were obtained. Polyurethanes with lower hard segment content (25-35%) were mechanically weak, but those having higher hard block content (45-60%) were rigid. The hard block content used for polymers reported herein was typically 40 wt %. IPDI-BD(40)/P(FOx/BrOx-m:n)
Table 2. Molecular Weights of Precursor and 5,5-Dimethylhydantoin-Substituted Polyurethanes by GPC after 5,5-dimethylhydantoin substitution
before 5,5-dimethylhydantoin substitution
a
precursor PU soft segment compositiona
Mn (× 103 g/mol)
Mw (× 103 g/mol)
Mn (× 103 g/mol)
Mw (× 103 g/mol)
P(3FOx/BrOx-2:1) P(3FOx/BrOx-1:1) P(3FOx/BrOx-1:2) P(5FOx/BrOx-2:1) P(5FOx/BrOx-1:1) P(5FOx/BrOx-1:2)
19 18 17 19 16 17
46 37 34 40 29 34
15 18 15 19 13 25
36 39 27 33 25 38
The hard block is IPDI/BD(40%).
Contraphilic Wetting for Polyurethanes
Langmuir, Vol. 23, No. 1, 2007 213
Table 3. Soft Block Tg’s before and after 5,5-Dimethylhydantoin Substitution polyurethane
Tg before (°C)
Tg after (°C)
PU-1 PU-2 PU-3 PU-4 PU-5 PU-6
-29 -29 -24 -27 -25 -29
11 17 2 10 13 16
conventional DSC. The soft block Tg was readily determined by conventional DSC, and MDSC was employed to detect the hard segment Tg. The DSC scans were performed from -70 to 150170 °C. We reported Tg’s previously for precursor (-CH2Br-containing) polyurethanes.17 The degree of phase separation in these systems was estimated using the glass-transition temperatures of the pure segments, the corresponding values in the polyurethane structure, and the Fox equation.17 Although detecting Tg’s for precursor polyurethanes was possible (but difficult because of broad inflections), we were not able to detect a hard segment Tg for the substituted polyurethanes using either conventional or MDSC. As indicated by the soft block Tg’s discussed below, there is substantial phase mixing promoted by hydrogen bonding involving the hydantoin moiety. Soft block Tg’s before and after 5,5-dimethylhydantoin substitution are shown in Table 3. The soft block Tg for precursor polyurethanes generally increased after 5,5-dimethylhydantoin substitution (∆Tg’s from 29 to 46 °C). For example, the soft block Tg for IPDI-BD(40)/P(3FOx/BrOx-1:1) (-29 °C) increased 46° after 5,5-dimethylhydantoin substitution (17 °C). The polar nature of the heterocyclic 5,5-dimethylhydantoin was responsible for the increase with the concomitant phase mixing noted above. PU-3 seems to be an outlier to the cited trend with a Tg of 2 °C, but the inflections due to the change in heat capacity are very broad as a result of phase mixing, resulting in low accuracy in the determination of Tg. Wetting Behavior. A dynamic contact angle analyzer (DCA) was employed following the Wilhelmy plate method.18,22,23 Previously, this method has facilitated the measurement of intrinsic contact angles for model PDMS networks24 and the elucidation of a copolymer effect for polyurethanes with semifluorinated soft blocks.21 Several advantages for the Wilhelmy plate method have been noted.25 A large polymer surface is interrogated in the DCA experiment compared to that in the sessile drop method.24 Another feature is the facile testing of water surface tension after sample analysis.24 This important control is performed using a glass coverslip (flamed to remove organic contaminants) to determine the water surface tension before and after evaluating a polymer coating. A decrease in post-use water surface tension indicates water surface contamination by surface-active materials such as oils and/or amphiphiles leached from the coating. The qualitative detection of water-
insoluble leached species is of practical importance for biomedical, electronic, and space applications. To study wetting behavior of the hydantoin-polyurethanes, glass coverslips were dip coated in 15 to 20 wt % THF solutions. After air drying for 2 h, the coated slides were placed in a vacuum oven at 50 °C for 24 h. Finally, all coatings were annealed for 24 h under vacuum at 85 °C. The resulting coatings were transparent, although some of the 5,5-dimethylhydantoincontaining coatings had a slightly yellow color. Table 4 lists polyurethane soft block compositions and advancing (θadv) and receding (θrec) contact angles for each immersion/emersion cycle (three or five). In a check of post-use water, no water contamination was observed for any coatings. The difference between the second and first advancing contact angle (∆C ) θadv2 - θadv1) defines the magnitude of the contraphilic effect. We previously reported the contraphilic effect for PU-4 surfaces with θadv ) 83° for the dry and 110° for the wetted surface (∆C ) 27°).14 We now report that after annealing at 85 °C under vacuum overnight, θadv was 68° for the dry and 106° for the wetted surface (∆C ) 38°, Table 4). The contraphilic effect was perfectly reversible upon drying the coating at 50 °C in vacuo. The observation of the amplification of ∆C on annealing at 85 °C under vacuum overnight for PU-4 led to the use of this protocol for all hydantoin-substituted polyurethanes (Table 4). Figure 2 shows DCA force versus distance curves (fdc’s) and goniometer drop profiles for two FOx/HyOx polyurethane coatings with high ∆C’s. Figure 2A shows fdc’s for PU-2 (∆C ) 34°), IPDI-BD(40)/P(3FOx/HyOx/BrOx-1.0:0.65:0.35), and Figure 2B shows fdc’s for PU-4 (∆C ) 38°), IPDI-BD(40)/ P(5FOx/HyOx/BrOx-2.0:0.7:0.3). The initial force value for the first fdc is positiVe (apparent mass gain) for both PU-2 and PU-4, indicating hydrophilic character as water “wicks” up onto the coated surface. The thermodynamic driving force is the gain in enthalpy through the formation of hydrogen bonds. After the first immersion/emersion cycle, the coated slide has been wetted to the maximum immersion line. During the second immersion cycle, an initial negatiVe force value (apparent mass loss) is obtained, indicating a hydrophobic surface. Now, waterwater hydrogen bonding interactions are stronger than watersurface interactions, and water resists wetting the surface. In addition to Wilhelmy plate analysis, the static contact angles and drop profiles for these polyurethanes are shown in Figure 2. The contact angles measured by drop profile analysis are close to those observed for DCA θadv. By observing the shape change of the actual drop, the switching of the surface from hydrophilic to hydrophobic with water is seen. Interestingly, water makes it hydrophobic. Table 4 shows θadv and θrec for the three compositions in the 3FOx and 5FOx series. The ratio of precursor FOx/BrOx polyurethanes was varied to provide high (2:1) and low levels of semifluorinated groups. Of particular note, within the 5FOx series, the polyurethane having the highest semifluorinated to
Table 4. Soft Block Composition and Advancing and Receding Contact Anglesa for 5,5-Dimethylhydantoin-Substituted Polyurethanesb soft block compositions
cycle1
cycle2
cycle3
pre-FOx/Br
after substitution
designation
θadv
θrec
θadv
θrec
θadv
θrec
∆Cd
2:1 1:1 1:2 2:1 1:1 1:2
P(3FOx/HyOx/BrOx-2.0:0.60:0.40) P(3FOx/HyOx/BrOx-1.0:0.65:0.35) P(3FOx/HyOx/BrOx-1.0:0.9:1.1) P(5FOx/HyOx/BrOx-2.0:0.7:0.3) P(5FOx/HyOx/BrOx-1.0:0.60:0.40) P(5FOx/HyOx/BrOx-1.0:1.1:0.9)
PU-1 PU-2c PU-3 PU-4 PU-5 PU-6c
72 70 93 68 83 93
32 35 28 38 38 42
101 104 104 106 108 105
34 37 34 40 40 42
100 101 104 106 107 104
34 38 36 41 41 42
29 34 11 38 25 12
a Estimated standard deviation (2°. b All coatings were annealed at 85 °C under vacuum overnight. c PU-2 and PU-6 were run for two additional cycles; θadv and θrec were identical to cycle 3. d ∆C ) θadv2 - θadv1.
214 Langmuir, Vol. 23, No. 1, 2007
Makal et al.
Figure 2. DCA force vs distance curves and goniometer drop profiles for (A) PU-2, IPDI-BD(40)/P(3FOx/HyOx/BrOx-1.0:0.65:0.35) and (B) PU-4, IPDI-BD(40)/P(5FOx/HyOx/BrOx-2.0:0.7:0.3).
5,5-dimethylhydantoin ratio (PU-4, ∼2.9) had the lowest initial contact angle (of 68°) when dry. The contraphilic effect was maximized for this composition (∆C ) 38°). PU-6, derived from a 5FOx/BrOx-1:2 polyurethane, has the lowest semifluorinated/5,5-dimethylhydantoin ratio (∼0.9). Dry PU-6 had θadv ) 93° and ∆C ) 12°. Originating from a 5FOx/ BrOx-1:1 polyurethane, PU-5 has an intermediate semifluorinated/5,5-dimethylhydantoin (∼1.7) ratio. The dry PU-5 surface had an intermediate θadv of 83° with a ∆C of 25°. Overall, the findings for the 5FOx series were counterintuitive because the polyurethane with the highest semifluorinated to 5,5-dimethylhydantoin ratio (PU-4, ∼2.9) was the most hydrophilic when dry (θadv of 68°). A different compositional dependence was found for the 3FOx series. The semifluorinated/5,5-dimethylhydantoin ratios for PU-1 (∼3.3) and PU-2 (∼1.5) are compositionally quite different, being derived from 3FOx/BrOx precursors with 2:1 and 1:1 ratios, respectively. However, PU-1 (72°) and PU-2 (70°) have similar initial θadv values. The contraphilic effect for these coatings was somewhat greater for PU-2 (34°) than for PU-1 (29°). PU-3 with a semifluorinated/5,5-dimethylhydantoin ratio of 1.1 provided quite different results. The dry PU-3 coating has a θadv of 93° and a ∆C of 12°. For the 3FOx series, the order of ∆C is PU-1 ≈ PU-2 > PU-3 whereas for the 5FOx series it is PU-4 > PU-5 > PU-6. The overall trend is that polyurethanes with higher semifluorinated/ 5,5-dimethylhydantoin ratios had surfaces that were more hydrophilic when dry. To examine whether contraphilic behavior would be reflected in wetting by an organic medium, PU-4 was examined using DCA in hexadecane (Figure 3). The hexadecane advancing contact angle for the dry surface (36°) is less than that for the wet surface (46°). The receding contact angle for the dry surface (26°) was less than that for the wet surface (30°). The hexadecane DCA (21) Makal, U.; Fujiwara, T.; Cooke, R. S.; Wynne, K. J. Langmuir 2005, 21, 10749-10755. (22) Hogt, A. H.; Gregonis, D. E.; Andrade, J. D.; Kim, S. W.; Dankert, J.; Feijen, J. J. Colloid Interface Sci. 1985, 106, 289-298. (23) Adamson, A. W.; Gast, A. P. Physical Chemistry of Surfaces, 6th ed.; John Wiley and Sons: New York, 1997. (24) Uilk, J. M.; Mera, A. E.; Fox, R. B.; Wynne, K. J. Macromolecules 2003, 36, 3689-3694. (25) Lander, L. M.; Siewerski, L. M.; Brittain, W. J.; Vogler, E. A. Langmuir 1993, 9, 2237.
Figure 3. Advancing and receding hexadecane contact angles for PU-6: (A) dry surface; (B) wetted surface
measurements confirm that the wetted coating is more oleophobic than when dry. However, the contact angles for either the dry or wet surface are moderate. For example, acrylates modified with semifluorinated comonomers have advancing contact angles up to 63°26 whereas perfluorinated polymers have values around 70°.27 The lower values for PU-4 are consistent with incomplete near-surface coverage of fluorinated moieties and the lack of order in the soft block. PU-4 coatings were investigated with other solvents (alcohols, methylene iodide), but swelling or dissolution occurred. The presence of amide groups, such as those present in 5,5dimethylhydantoin, have previously been shown to produce dramatic surface phenomena. Intramolecular hydrogen bonding (above LCST) and hydrogen bonding with water (below LCST) are responsible for hydrophobic (above LCST)/hydrophilic (below LCST) switching for a poly(N-isopropylacrylamide) surface.2 Like the amphiphilic response, we believe that contraphilic wetting is driven enthalpically. Shown in Figure 4 is the proposed mechanism for the contraphilic wetting. For the dry surface (4A), enthalpically driven hydrogen bonding of 5,5-dimethylhydantoin (26) Thomas, R. R.; Anton, D. R.; Graham, W. F.; Darmon, M. J.; Stika, K. M. Macromolecules 1998, 31, 4595-4604. (27) Tavana, H.; Appelhans, D.; Zhuang, R. C.; Zschoche, S.; Grundke, K.; Hair, M. L.; Neumann, A. W. Colloid Polym. Sci. 2006, 284, 497-505.
Contraphilic Wetting for Polyurethanes
Langmuir, Vol. 23, No. 1, 2007 215
Figure 4. Proposed mechanism for contraphilic wetting: (4A) hydrophilic dry surface; (4B) hydrophobic wet surface.
amide groups to acidic methylene hydrogens of semifluorinated groups disrupts the normal surface concentration of semifluorinated groups whereas the geometric arrangement of hydantoin N-H (4A) results in availability for hydrogen bonding with water. This accounts for the hydrophilic surface of the dry state. The diad sequence shown in Figure 4 would be more probable in compositions having higher ratios of semifluorinated/5,5dimethylhydantoin side chains. This proposed mechanism agrees with the observation of higher ∆C’s for soft blocks with higher FOx/HyOx ratios. The proposed mechanism for hydrogen bonding in the “dry, hydrophilic state” involves carbonyl hydrogen bonding to a CH2 group adjacent to two strongly electron-withdrawing groups (viz., -O- and -CF2-). This interaction finds precedence in the well-known hydrogen bonding of the carbonyl group in polymethyl methacrylate (PMMA) to the -CH2- group in polyvinylidene fluoride (PVDF). Several vibrational spectroscopic studies demonstrate that this specific interaction is the source of the enthalpically driven miscibility of PVDF and PMMA.28-30 The participation of the -CF2- group in an interaction involved in contraphilic switching is precluded by the observation that 3FOx/HyOx containing soft blocks shows an extent of contraphilic behavior that is similar to those containing 5FOx/HyOx segments. Upon exposure to water (Figure 4, 4B), amide groups switch from hydrogen bonding to -CH2CF2CF3 to stronger hydrogen bonding with water. As a result, semifluorinated groups are “released”, and the surface becomes hydrophobic. Upon dehydration of the surface, the initial hydrophilic wetting character is restored. This cycling from the dry hydrophilic state to the wet hydrophobic state may be repeated many times. If the coating is left at ambient humidity, then θadv has a value that is intermediate between those of the dry and wet states. After the coating was left overnight (∼30% relative humidity), the advancing contact angle for PU-4 is ∼93°. The detection limit of measuring water uptake for the polyurethane coatings was considered in order to gain an understanding of the extent of water interaction. The time of coating immersion during a DCA cycle is 3 min. Three to five cycles were run so that the maximum time of immersion was about 15 min. The mass of polymer coating on a coverslip is typically about 50 mg whereas the accuracy of the electrobalance is (0.1 mg. The minimum detectable gain in mass is therefore 0.2%. Within experimental error, no sign of mass gain was observed after removing the coating from water after all immersion/emersion cycles (Figure 2). On the basis of these (28) Leonard, C.; Halary, J. L.; Monnerie, L. Polymer 1985, 26, 1507-1513. (29) Leonard, C.; Halary, J. L.; Monnerie, L. Macromolecules 1988, 21, 29882994. (30) Kim, K. J.; Cho, Y. J.; Kim, Y. H. Vib. Spectrosc. 1995, 9, 147-159.
Figure 5. Conversion of N-amide to N-chloramide with hypochloride and reduction of N-chloramide to N-amide with thiosulfate.
observations, the water interaction that results in the switch from hydrophilic to hydrophobic wetting must be an adsorption/nearsurface phenomenon. Contraphilic-like wetting behavior has been reported previously. Ferguson has reported that oxidized 1,2-polybutadiene surfaces became more hydrophobic when exposed to hot water.4,5 This is an entropically driven process. The entropic loss due to the stretched chains translates into the elastic restoring force as the temperature increases. Thus, the polar groups are pulled away from the water-polymer interface, gradually increasing the hydrophobicity of the surface, but the change in the wetting behavior of the oxidized 1,2-polybutadine surface was damped after successive cycles. In contrast, the contraphilic wetting is not attenuated by repeated cycling between dry and wet states. According to the proposed mechanism, if the hydrogen bonding capability of the hydantoin moiety is removed, then the contraphilic effect should be attenuated or should disappear. To decrease the surface amide concentration, a PU-4 coating was subjected to treatment with hypochlorite for 1 h.15 This treatment effects the conversion of near-surface amide 5A to chloramide 5B according to Figure 5. This reaction has received considerable study because chloramide form 5B lends oxidative antimicrobial character to the surface.31,32 The wetting behavior of the resulting chloramide-functionalized coating PU-4 was investigated by DCA analysis. The first cycle gave θadv1 ) 101.3 ( 1.7° and θrec1 ) 41.6 ( 0.6°. This high advancing/low receding contact angle behavior is characteristic of polymers with side-chain semifluorinated groups.33-37 A (31) Lin, J.; Winkelmann, C.; Worley, S. D.; Kim, J. H.; Wei, C. I.; Cho, U. C.; Broughton, R. M.; Santiago, J. I.; Williams, J. F. J. Appl. Polym. Sci. 2002, 85, 177-182. (32) Sun, Y.; Sun, G. J. Appl. Polym. Sci. 2002, 84, 1592-1599. (33) Katano, Y.; Tomono, H.; Nakajima, T. Macromolecules 1994, 27, 23422344. (34) Wang, J. G.; Mao, G. P.; Ober, C. K.; Kramer, E. J. Macromolecules 1997, 30, 1906-1914. (35) Xiang, M. L.; Li, X. F.; Ober, C. K.; Char, K.; Genzer, J.; Sivaniah, E.; Kramer, E. J.; Fischer, D. A. Macromolecules 2000, 33, 6106-6119. (36) Johnston, E.; Bullock, S.; Uilk, J.; Gatenholm, P.; Wynne, K. J. Macromolecules 1999, 32, 8173-8182. (37) Bertolucci, M.; Galli, G.; Chellini, E.; Wynne, K. J.; Uilk, J. Polym. Prepr. 2002, 43, 344.
216 Langmuir, Vol. 23, No. 1, 2007
second DCA cycle and successive ones showed slightly different results (θadv2 ) 104.5 ( 0.6°, θrec2 ) 43.5 ( 1.3°), suggesting that conversion to chloramide was not quite complete, but the conversion of amide to chloramide virtually eliminates the contraphilic effect. Immersing the PU-4 coating in thiosulfate for 1 h reduces chloramide 5B back to amide 5A (Figure 5). Contraphilic wetting behavior is restored and is indistinguishable from the pristine coating. Cycling between amide and chloramide as shown in Figure 5 may be repeated at least four times with identical results.
Conclusions In water, surface reorganization for polar polymers has a strong thermodynamic driving force. Hydrogen bonding of near-surface polar groups with water is driven enthalpically, easily overcoming unfavorable entropy in forming a more ordered molecular arrangement. Upon exposure to water, soft surfaces rearrange opening paths to more polar hydrophilic moieties so that hydrophilicity often increases with time.38-41 The migration of hydrophilic polar groups to the polymer surface and water adsorption or absorption gradually decrease the advancing contact angle. What we have observed is the exact opposite of anticipated wetting behavior: the surfaces of the contraphilic polyurethanes become more hydrophobic when wet and upon dehydration return to their original hydrophilic character. The contraphilic effect was found serendipitously while evaluating the wetting behavior of polyurethanes having soft (38) Andrade, J. D.; Tingey, K. G. Langmuir 1991, 7, 2471-2478. (39) Senshu, K.; Yamashita, S.; Mori, H.; Ito, M.; Hirao, A.; Nakahama, S. Langmuir 1999, 15, 1754-1762. (40) Pike, J. K.; Ho, T.; Wynne, K. J. Chem. Mater. 1996, 8, 856-860. (41) Lemieux, M.; Usov, D.; Minko, S.; Stamm, M.; Shulha, H.; Tsukruk, V. V. Macromolecules 2003, 36, 7244-7255.
Makal et al.
blocks containing 5,5-dimethylhydantoin and semifluorinated side chains.15 Like conventional amphiphilic wetting, we propose that contraphilic wetting is driven enthalpically by hydrogen bonding (Figure 4). The model proposed in Figure 4 leaves several issues unaddressed. How does the amide function determine dry wetting behavior? Would it not be expected that the 5,5-dimethyl groups would have a strong influence on dry wetting behavior? The 5FOx chain extends six atoms from the main chain, but there are only five atoms from the main chain to the amide H. How can the shorter chain determine dry wetting behavior? A combination of modeling and better defined chain composition may answer these questions in the future. The application of vibrational spectroscopic techniques will be challenging because of bulk and near-surface carbonyl and hydrocarbon absorptions. The reaction-on-polymer approach to obtaining the 5,5hydantoin-substituted soft blocks was the only tractable method to obtain the desired soft block compositions. This method has several unfavorable characteristics. Considering that (1) the precursor polyurethane P(FOx/BrOx) soft blocks are derived from relatively low molecular weight telechelics, (2) the reactionon-polymer method produces a soft block with residual bromomethyl groups and a chain architecture of unknown sequence distribution, and (3) after substitution, the soft blocks have Tg’s barely below ambient (Table 3), we conclude that the contraphilic effect is quite robust. Future work will be aimed at a simpler preparative method for creating soft block structures with better defined compositions and chain architectures. Acknowledgment. We are grateful to the National Science Foundation for support (DMR-0207560). We thank Dr. Steven Aubuchon, TA Instruments for thermal analysis technical advice and equipment. LA0615600
