Langmuir 2003, 19, 2533-2535
2533
Letters General Approach for the Preparation of Nanoscale Inorganic Layers on Polymeric Materials Surfaces Mamle Quarmyne and Wei Chen* Chemistry Department, Mount Holyoke College, South Hadley, Massachusetts 01075 Received November 4, 2002 A general approach for the preparation of robust nanoscale inorganic layers on polymer film substrates is presented. Poly(tetrafluoroethylene-co-hexafluoropropylene), poly(4-methyl-1-pentene), and poly(ethylene terephthalate) were “activated” by the spontaneous adsorption of poly(vinyl alcohol); silica and titania were subsequently condensed on these surfaces by sequential vapor phase reaction with either silicon tetrachloride or titanium tetrachloride and water. The thickness and wettability of the inorganic layers can be controlled by the number of SiCl4 or TiCl4 and H2O reaction cycles. Reactions between the nanoscale inorganic layers and monochlorosilanes for silica and hydridosilanes for titania indicate that these surfaces have silica-like and titania-like reactivities.
Organic-inorganic hybrids have been prepared to take advantage of the desirable properties of each class of materials. Most of the literature concerns the preparation of bulk organic/inorganic composites.1-5 Our objective is to develop a general approach for fabrication of nanoscale inorganic surface layers on organic materials. Polymeric materials offer obvious processing advantages, but they often fall short concerning surface properties. Inorganics are much more easily surface-modified for a range of applications, such as catalysis, adhesion, wettability, biocompatibility, sensing, and protective coating. There is one report6 of the surface modification of poly(ethylene terephthalate) with 3-aminopropyltrialkoxysilanes to prepare silica-like surfaces; however, the approach is specific to polyesters and amine-containing silanes and cannot be generalized to other systems. We report the preparation of silica-like and titania-like surfaces on three polymer film samples, poly(ethylene terephthalate) (PET), poly(4-methyl-1-pentene) (PMP), and poly(tetrafluoroethylene-co-hexafluoropropylene) (FEP). Silica and titania were chosen because they are technologically important inorganic materials; polyesters, polyolefins, and fluoropolymers are commercially important polymers that represent a wide range of polymers with different chemical compositions and surface energetics. These systems were chosen to illustrate the generality of the approach shown in Scheme 1. Polymers are “primed” by adsorption of poly(vinyl alcohol) (PVOH) from aqueous solution to solid/liquid interfaces to introduce reactive -OH groups. This adsorption (modification) is general for hydrophobic materials. Cross-linking of PVOH with glutaraldehyde was carried out to prevent the adsorbed PVOH layer from desorption in organic solvents and high* To whom correspondence should be addressed. E-mail:
[email protected]. Tel: 413-538-2224. Fax: 413-538-2327. (1) Kresge, C. T.; Leonowicz, M. E.; Roth, W. J.; Vartuli, J. C.; Beck, J. S. Nature 1992, 359, 710. (2) Davis, M. E.; Lobo, R. F. Chem. Mater. 1992, 4, 756. (3) Cheetham, A. K.; Fe´rey, G.; Loiseau, T. Angew. Chem., Int. Ed. 1999, 38, 3268. (4) Puntes, V. F.; Krishnan, K. M.; Alivisatos, A. P. Science 2001, 291, 2115. (5) Imhof, A.; Pine, D. J. Nature 1997, 389, 948. (6) Fadeev, A. Y.; McCarthy, T. J. Langmuir 1998, 14, 5586.
Scheme 1
temperature aqueous media. Nanoscale silica and titania layers are prepared via reactions (multiple cycles if necessary) between -OH groups (alcohol, silanol, or titanol) with SiCl4 or TiCl4 followed by hydrolysis (air exposure). These silica and titania layers react as inorganics, as assessed by their reactivities toward several silanes. We recently reported7-9 the unique adsorption behavior of PVOH at aqueous solution/hydrophobic solid interfaces. The adsorption is driven by crystallization of PVOH chains at the solid/solution interface. We demonstrated that this process is a general approach to hydrophilizing and chemically activating inert surfaces. Using the optimal conditions reported,8 PVOH was adsorbed from aqueous solution for 24 h at room temperature to PET, PMP, and FEP film samples; stabilization of the adsorbed PVOH was achieved by cross-linking the polymer using an acidic aqueous solution of glutaraldehyde.8 Reactions between the alcohol-containing polymer substrates and either SiCl4 or TiCl4 were carried out in the vapor phase. Sequential reactions between SiCl4 or TiCl4 and water (present in air) were carried out. The effects of reaction kinetics and the number of reaction cycles on the thickness and wettability of the inorganic layers were examined. Reactivities of the resulting silica-like and titania-like surfaces were assessed with reactions with monochlorosilanes and hydridosilanes, respectively. All samples studied were characterized by contact angle and X-ray photoelectron spectroscopy (XPS). (7) Coupe, B.; Chen, W. Macromolecules 2001, 34, 1533. (8) Kozlov, M.; Quarmyne, M.; Chen, W.; McCarthy, T. J. Macromolecules 2003, 36, in press. (9) Chen, W. U.S. Patent filed.
10.1021/la0208880 CCC: $25.00 © 2003 American Chemical Society Published on Web 03/07/2003
2534
Langmuir, Vol. 19, No. 7, 2003
Letters
Figure 1. Silicon content obtained from XPS atomic composition data at a 75° takeoff angle on cross-linked samples of PETPVOH (b), PMP-PVOH (O), and FEP-PVOH (0) as a function of the number of SiCl4 reaction cycles. Table 1. Advancing and Receding Water and Hexadecane Contact Angles (Degrees) and XPS Atomic Composition (75° Takeoff Angle) for Cross-linked PET-PVOH and Its Derivatives substrates
H2O (θA/θR)
C16H34 (θA/θR)
C%
O%
Si/Ti/F%
controla -SiO2b -SiO2-CF3c -SiO2-CH3c -SiO2-CO2CH3c -TiO2d -TiO2-CH3e
58/14 20/7 109/83 99/90 79/75 26//7 103/90
spreads spreads 72/49 spreads spreads spreads 38/32
73.32 30.56 31.53 39.48 47.73 31.54 49.28
26.68 54.82 25.82 46.20 41.78 49.11 36.60
14.61Si 8.90Si 33.75F 14.32Si 10.49Si 19.35Ti 12.31Ti 1.81Si
a Cross-linked PET-PVOH. b Obtained after 10 reaction cycles with SiCl4. c Substrates were exposed to various silane vapors at 50 °C for 114 h. d Obtained after 7 reaction cycles with TiCl4. e Reaction in toluene solution at 50 °C for 79 h.
A kinetics study indicated that the reaction between cross-linked PET-PVOH and SiCl4 was complete after 15 min; this reaction time was used for all PVOHcontaining polymer substrates and SiCl4. Figure 1 indicates that silicon content, determined by XPS for each of the three substrates, increased steadily as a function of the number of reaction cycles. The rate of silica growth is greater on cross-linked PET-PVOH than on the other two substrates as indicated by the differences in slopes. This trend is consistent with the observation of lower water contact angles, θA/θR, observed for PET-PVOH (40°/13°) relative to PMP-PVOH (58°/16°) and FEP-PVOH (63°/ 18°).8 The more hydrophilic nature of PET-PVOH due to the availability of more free surface -OH groups allows more SiCl4 to react and thus more silica to grow during each cycle. Since there is no other difference among the three polymer supports, data on PET only are presented in the remainder of this manuscript. As the number of reaction cycles increases, water contact angles of polymersupported silica decrease steadily and become similar to those of clean silicon wafers. The contact angle and atomic composition data of cross-linked PET-PVOH before and after 10 cycles of reaction with SiCl4 are given as the first and second entries in Table 1. XPS data indicate that the total thickness of the silica layer after 10 cycles of reaction is less than ∼40 Å, the sampling depth of XPS at a 75° takeoff angle.10 This is a lower estimate due to the issue of carbon contamination of high-energy silica surfaces. Atomic force microscope images (not shown) of the surfaces during silica growth indicate that surface topography and thus surface roughness remain the same. The reactivity
Figure 2. Silicon and titanium content obtained from XPS atomic composition data at a 75° takeoff angle on cross-linked PET-PVOH as a function of the number of SiCl4 (b) and TiCl4 (O) reaction cycles, respectively.
of the nanoscale-thick silica layer was assessed by reactions with n-decyldimethylchlorosilane, tridecafluoro1,1,2,2-tetrahydrooctyl-dimethylchlorosilane, and 10-(carbomethoxy)decyldimethylchlorosilane to form -CH3-, -CF3-, and -CO2CH3-containing surfaces. Water contact angles increase dramatically after all three silane reactions and silicon contents decrease, indicating the presence of covalently attached hydrophobic groups on silica surfaces. Comparing the hexadecane contact angles of 72°/ 49° on the PET-PVOH-SiO2-CF3 surface with the literature θA values11 of 45° for poly(tetrafluoroethylene) and 72° for a monolayer containing -CF3 groups at the surface indicates that the perfluoroalkyl chains lie at some angle to the surface exposing primarily -CF3 end groups. Both the atomic composition and contact angle data before and after the silane reactions, as shown in Table 1, indicate that the nanoscale-thick silica layers possess silica-like reactivity. Reactions between alcohol-containing polymer substrates and TiCl4 take place faster than those with SiCl4. TiCl4 is a stronger Lewis acid, and this is the likely reason for this difference; 5 min was found to be sufficient for each reaction cycle. The growth rate of titania is faster than that of silica as indicated by the greater slope of the plot of titanium content versus the number of reaction cycles shown in Figure 2 for cross-linked PET-PVOH surfaces. The likely explanation for this is that TiCl4 reacts monofunctionally more than does SiCl4, which is supported by the fact that hydrolysis of SiCl4 occurs completely while only partial hydrolysis of TiCl4 takes place in the absence of a base.12 As the number of reaction cycles increases, water contact angles of polymer-supported titania decrease steadily. The contact angle and atomic composition data after seven cycles of reaction with TiCl4 are given as the sixth entry in Table 1. An accurate thickness assessment of the titania layer could not be made using XPS data, due to the issue of carbon contamination of high-energy titania surfaces, but an estimate of several nanometers is on scale.10 (10) This estimate is supported by the attenuation of the fluorine content of cross-linked FEP-PVOH samples measured at a 75° takeoff angle as the number of reaction cycles increases: it decreases from 29.86% to 15.48% after 10 reaction cycles with SiCl4 and to 12.98% after 7 reaction cycles with TiCl4. (11) Zisman, W. A. Adv. Chem. Ser. 1964, 43, 1. (12) Cotton, F. A.; Wilkinson, G. Advanced Inorganic Chemistry, A Comprehensive Text, 4th ed.; John Wiley & Sons: New York, 1980; pp 385 and 694.
Letters
Langmuir, Vol. 19, No. 7, 2003 2535
The reactivity of the nanoscale-thick titania layer was assessed by reactions with hydridosilanes.13 The result of one reaction is shown in Table 1. After reaction with octadecylsilane, a decrease in titanium content, the appearance of a silicon peak in the XPS spectrum, and a dramatic increase in hydrophobicity (water contact angles) and lyophobicity (hexadecane contact angles) indicate the covalent attachment of a close-packed octadecyl-containing monolayer to the surface of the nanoscale-thick titania layer. Reactions using other inorganic reagents to fabricate inorganic layers on other smooth as well as high surface area supports are in the process of being evaluated. We view this approach as a general yet powerful method to yield materials that could be important in many applications. Experimental Procedure General. PVOH (MW ) 108 000), 99.7 mol % hydrolyzed, was obtained from Polysciences. FEP film (5 mil) was obtained from Berghof, PET film (5 mil Mylar) was obtained from duPont, and PMP film (2 mil) was obtained from Mitsui. House-purified water (reverse osmosis) was further purified using a Millipore Milli-Q system that involves reverse osmosis, ion-exchange, and filtration steps. Other reagents and solvents were obtained from Aldrich or Fisher and used as received. X-ray photoelectron spectra were recorded with a Perkin-Elmer-Physical Electronics 5100 with Mg KR excitation (400 W). Spectra were obtained at 15° and 75° takeoff angles (between the plane of the surface and the entrance lens of the detector optics), which give the composition of the outermost ∼10 and ∼40 Å, respectively. Contact angle measurements were made with a Rame´-Hart telescopic goniometer and a Gilmont syringe with a 24-gauge flat-tipped needle. The probe fluid used was water, purified as described above. Dynamic advancing (θA) and receding angles (θR) were recorded while the probe fluid was added to and withdrawn from the drop, respectively. Contact angle assesses the composition of the outer few angstroms. Polymer Substrates. FEP film was cleaned by immersing the film samples in 25 mL of concentrated sulfuric acid containing (13) Fadeev, A.; McCarthy, T. J. J. Am. Chem. Soc. 1999, 121, 12184.
0.5 g of potassium chlorate for 2 h. The film samples were then rinsed with 5 aliquots of water and then THF and dried at reduced pressure overnight. PET film was rinsed with water and then methanol, extracted with hexane for 2 h, and dried at reduced pressure overnight. PMP film samples were extracted with refluxing dichloromethane for 1 h and then dried at reduced pressure overnight. Adsorption of PVOH. PVOH was dissolved in Milli-Q water in a glass vial for approximately 1 h while stirring and heating to ∼100 °C. The solution was left to cool to room temperature and then made to 0.01 M (based on repeat units). A polymeric film sample, FEP, PMP, or PET, was immersed in the PVOH aqueous solution for 24 h at room temperature. The film sample was then rinsed with 3 aliquots of Milli-Q water and dried at reduced pressure overnight. Cross-Linking of PVOH. A PVOH adsorbed film sample was submerged in a 0.075 M aqueous solution of glutaraldehyde containing 0.2 M sulfuric acid for 30 min at 40 °C. The film sample was rinsed with 3 aliquots of Milli-Q water and dried at reduced pressure overnight. Reactions with SiCl4/TiCl4. All samples were dried in a Schlenk flask for 2 h under reduced pressure and purged with N2. The samples were exposed to SiCl4 vapor via a cannula connecting the reaction flask and a SiCl4 reservoir in a steady stream of N2 at room temperature for 15 min. At the end of the reaction, residual SiCl4 in the flask was purged with N2 before the samples were exposed to air for 5 min to allow hydrolysis to take place. Before the next cycle, the sample flask was pumped down under a vacuum and filled with N2. Reactions with TiCl4 were carried out in a similar fashion as with SiCl4 except that reaction time for each cycle is 5 min and samples were vacuumdried for 2 h between cycles to remove the reaction side product, HCl.
Acknowledgment. We thank the National Science Foundation and the Petroleum Research Fund for financial support and the NSF-sponsored Materials Research Science and Engineering Center (MRSEC) for use of the central facilities at the University of Massachusetts at Amherst. LA0208880
