9392
Langmuir 2008, 24, 9392-9400
Immobilization of a Hyperbranched Polyester via Grafting-to and Electron Beam Irradiation Senta Reichelt, Uwe Gohs, Frank Simon, Sven Fleischmann, Klaus-Jochen Eichhorn,* and Brigitte Voit Leibniz Institute of Polymer Research Dresden e.V., Hohe Strasse 6, 01069 Dresden, Germany ReceiVed March 3, 2008. ReVised Manuscript ReceiVed May 19, 2008 Stable thin films of an aromatic-aliphatic hyperbranched polyester with hydroxyl groups were fabricated on silicon substrates using electron beam irradiation and a grafting-to approach. We present a detailed study on the influence of the dose, dose rate, and temperature on the film properties and degradation behavior of the polyester immobilized by electron beam irradiation. A patterned polyester film was prepared on the substrate using a masking technique. In the second part of this work, we report on a method for the strong binding of the hyperbranched polyester onto the surface of an “activated” silicon substrate without using any coupling agent. The results are compared with the grafting-to of the hydroxyl-terminated polyester using thin PGMA anchoring layers (Reichelt et al. Macromol. Symp. 2007, 254, 240-247). The optimal conditions and mechanism of the anchoring procedures were investigated. The surface and film properties of all immobilized polymer films were characterized by atomic force microscopy (AFM), X-ray photoelectron spectroscopy (XPS), electrokinetic measurements, contact-angle measurements by drop-shape analysis, spectroscopic and imaging ellipsometry, and infrared spectroscopy. It is shown that all immobilization methods can be optimized in such a way that the polymer surface properties remain unchanged compared to those of nongrafted polyesters.
1. Introduction Since the beginning of the research on hyperbranched polymers (HBPs) in the early 1990s, a number of reviews concerning their synthesis, properties, and applications have been published.2–7 Already in 1952, Flory reported on the theory behind the polycondensation of ABx monomers, leading to hyperbranched architectures.8 In contrast to the well-defined dendrimers, hyperbranched (hb) polymers can be easily synthesized by onestep polycondensation or polyaddition, resulting in randomly branched macromolecules.2–7 Specific properties such as low solution viscosity and high solubility determined by the globular branched shape and the high number of functionalities are the basis for a variety of applications. Their utility has been shown, for instance, in medicinal chemistry,9 in microelectronics,10 as additives in linear polymers,11 and in coating technology.11–15 Thin films of HBPs are also promising materials for sensor * To whom correspondence should be addressed. E-mail:
[email protected]. Phone: +49-351-4658-256. Fax: +49-351-4658-565. (1) Reichelt, S.; Jakisch, L.; Simon, F.; Grundke, K.; Eichhorn, K.-J.; Voit, B. Macromol. Symp. 2007, 254, 240–247. (2) Voit, B. J. Polym. Sci. A: Polym. Chem. 2005, 43, 2679–2699. (3) Gao, C.; Yan, D. Prog. Polym. Sci. 2004, 29, 183–275. (4) Voit, B. I. C. R. Chim. 2003, 6, 821–832. (5) Jikei, M.; Kakimoto, M. Prog. Polym. Sci. 2001, 26, 1233–1285. (6) Hawker, C. J. AdV. Polym. Sci. 1999, 147, 143–160. (7) Voit, B. J. Polym. Sci. A: Polym. Chem. 2000, 38, 2505–2525. (8) Flory, P. J. J. Am. Chem. Soc. 1952, 74, 2718–2723. (9) Stiriba, S.-E.; Frey, H.; Haag, R. Angew. Chem., Int. Ed. 2002, 41, 1329– 1334. (10) Voit, B.; Eigner, M.; Estel, K.; Wenzel, C.; Bartha, J. W. Macromol. Symp. 2002, 177, 147–154. (11) Voit, B.; Beyerlein, D.; Eichhorn, K.-J.; Grundke, K.; Schmaljohann, D.; Loontjens, T. Chem. Eng. Technol. 2002, 25, 704–707. (12) Johansson, M.; Malmstro¨m, E.; Hult, A. J. Polym. Sci. A: Polym. Chem. 1993, 31, 619–624. (13) Sidorenko, A.; Zhai, X. W.; Peleshanko, S.; Greco, A.; Shevchenko, V. V.; Tsukruk, V. V. Langmuir 2001, 17, 5924–5934. (14) Sidorenko, A.; Zhai, X. W.; Greco, A.; Tsukruk, V. V. Langmuir 2002, 18, 3408–3412. (15) Sidorenko, A.; Zhai, X. W.; Simon, F.; Pleul, D.; Tsukruk, V. V. Macromolecules 2002, 35, 5131–5139.
applications.16–19 Aromatic hyperbranched polyesters with different functionalities (hydroxyl, carboxyl, acetyl) are sensitive to water vapor,20 alcohols,17,18 freons,17 and nerve agents, as well as explosive simulants.21 In previous works, we studied thin films of these hb polyesters with regard to swelling and protein adsorption behavior.19,22 Similarly, Siegers et al. described the self-assembly of thin films of dendritic poly(glycerol) (PG) that hinder protein adsorption.16 In contrast to the dendritic PGs, hyperbranched aromatic polyesters with hydroxyl functionalities adsorb lysozyme and human serum albumin.22 Mostly, thin films of hyperbranched polyesters have been prepared to date by spin coating on solid substrates from polymer solution without additional immobilization. As a result, only noncovalent interactions such as H-bonding, electrostatic, or weak van der Waals forces were responsible for the substrate layer adhesion. Nevertheless, many applications require more stable polymer films in liquid media (e.g., water, buffer solutions, organic solvents). The immobilization of polymer films has been the subject of intensive experimental and theoretical research.23 An appropriate method is the attachment of endfunctionalized polymers to a surface-modified substrate (graftingto method). Thus, silicon substrates were modified by Minko et al. with ultrathin functional films of epoxysilane [(3-glycidoxypropyl)trimethoxysilane, GPS] and by Zdyrko et al. with poly(glycidyl methacrylate) (PGMA), providing reactive epoxy groups for the anchoring of carboxy-terminated polystyrene (PS), (16) Siegers, C.; Biesalski, M.; Haag, R. Chem. Eur. J. 2004, 10, 2831–2838. (17) Belge, G.; Beyerlein, D.; Betsch, C.; Eichhorn, K.-J.; Gauglitz, G.; Grundke, K.; Voit, B. J. Anal. Bioanal. Chem. 2002, 374, 403–411. (18) Vollprecht, M.; Dieterle, F.; Busche, S.; Gauglitz, G.; Eichhorn, K.-J.; Voit, B. Anal. Chem. 2005, 77, 5542–5550. (19) Mikhailova, Y.; Dutschk, V.; Bellmann, C.; Grundke, K.; Eichhorn, K.J.; Voit, B. Colloids Surf. A: Physicochem. Eng. Aspects 2006, 279, 20–27. (20) Beyerlein, D.; Belge, G.; Eichhorn, K.-J.; Gauglitz, G.; Grundke, K.; Voit, B. Macromol. Symp. 2001, 164, 117–131. (21) Hartmann-Thompson, C.; Keeley, D. L.; Voit, B.; Eichhorn, K.-J.; Mikhailova, Y. J. Appl. Polym. Sci. 2008, 107, 1401–1406. (22) Mikhailova, Y.; Dutschk, V.; Mu¨ller, M.; Eichhorn, K.-J.; Voit, B. Colloids Surf. A: Physicochem. Eng. Aspects 2007, 297, 19–29. (23) Zhao, B.; Brittain, W. J. Prog. Polym. Sci. 2000, 25, 677–710.
10.1021/la8006662 CCC: $40.75 2008 American Chemical Society Published on Web 07/23/2008
Immobilization of a Hyperbranched Polyester Scheme 1. Synthetic Approach for Aromatic-Aliphatic Hyperbranched Polyester (aaPOH)
Langmuir, Vol. 24, No. 17, 2008 9393
Surface and film properties of aaPOH were analyzed combining different analytical methods: atomic force microscopy (AFM), drop-shape analysis (DSA), streaming potential measurements, X-ray photoelectron spectroscopy (XPS), Fourier transform infrared (FT-IR) spectroscopy, and ellipsometry. Products resulting from appropriate model experiments using the corresponding bulk material were also characterized by differential scanning calorimetry (DSC) and size-exclusion chromatography (SEC).
2. Experimental Section poly(vinylpyridine) (PVP), or poly(ethylene glycol) (PEG).24,25 Recently, Sidorenko et al. reported on the chemical reaction between surface groups and functionalized hyperbranched polymers. They commented on the heat-mediated melt graftingto of epoxy-functionalized aliphatic hyperbranched polymers to the hydroxyl groups of a bare silicon substrate14 using the known reaction between silanol and epoxy groups.26 In another work, Mikhailova et al. studied the grafting of carboxy-terminated hyperbranched aromatic polyesters on thin layers of GPS.27 In a previous work, we reported on the immobilization of a hyperbranched hydroxyl-terminated aromatic-aliphatic polyester using 1,3-phenylene bisoxazoline as the coupler.1 Apart from the grafting-to approach, recent work has focused on the synthesis of hydrogel films by electron beam irradiation utilizing a completely different approach for immobilization and cross-linking of vinyl-based polymers without the preformation of additional functional layers.28,29 The high-energy electron treatment results in the formation of macroradicals. Simultaneously, the stabilization of the film through the intermolecular reaction of radicals via cross-linking and anchoring to the silicon substrate is achieved. The switching behavior of electron-beamirradiated poly(vinyl pyrrolidone)- and poly(vinyl methyl ether)based hydrogels was studied by Meinhold et al.28 and Hegewald et al.29 Both hydrogel films persist through several swelling and deswelling processes without detachment. To our knowledge, there are no reports in the literature on the formation of homogeneous covalently immobilized functional hyperbranched polymer films in a thickness range from 10 to 100 nm, which is promising for sensor applications. In the course of this study, two approaches to the immobilization of a methylene group containing aromatic-aliphatic hyperbranched OHterminated polyester (aaPOH) were examined. Important goals are preserving the typical HBP globular architecture within the thin film as far as possible and obtaining a sufficient accessibility to a large number of functional HBP end groups for interactions with other molecules (analytes, proteins). In the first part, we focus on the preparation of stable films on silicon wafers via electron beam treatment. The investigation covers the influence of different experimental settings on the film thickness d and the degradation behavior. The second part discusses (1) the grafting of the polyester to bare silicon substrates with native silicon dioxide layers that are rich in silanol groups and (2) a comparison of the new immobilization method with previously obtained results on thin macromolecular anchoring layers of PGMA.1 (24) Minko, S.; Patil, S.; Datsyuk, V.; Simon, F.; Eichhorn, K.-J.; Motornov, M.; Usov, D.; Tokarev, I.; Stamm, M. Langmuir 2002, 18, 289–296. (25) Iyer, K. S.; Zdyrko, B.; Malz, H.; Pionteck, J.; Luzinov, I. Macromolecules 2003, 36, 6519–6526. (26) Hortschansky, P.; Heublein, G. Makromol. Chem. 1991, 192, 1535–1540. (27) Mikhailova, Y.; Pigorsch, E.; Grundke, K.; Eichhorn, K.-J.; Voit, B. Macromol. Symp. 2004, 210, 271–280. (28) Meinhold, D.; Schweiss, R.; Zschoche, S.; Janke, A.; Baier, A.; Simon, F.; Dorschner, H.; Werner, C. Langmuir 2004, 20, 396–401. (29) Hegewald, J.; Schmidt, T.; Gohs, U.; Gu¨nther, M.; Reichelt, R.; Stiller, B.; Arndt, K.-F. Langmuir 2005, 21, 6073–6080.
Materials. The hyperbranched OH-terminated polyester (aaPOH) 2 with an aromatic-aliphatic backbone (Scheme 1) was synthesized by melt polycondensation of 4,4-bis(4′-hydroxyphenyl) valeric acid 1 at 185 °C in a vacuum.30The Fre´chet-type degree of branching (DB) of 0.5 was calculated from the linear, dendritic, and terminal repeating units of the aromatic part separated by 13C NMR spectroscopy.31,32 A molecular weight of 18 000 g/mol [polydispersity index(PDI) of 2.8] was determined.33 The aaPOH was stable up to 300 °C (TGA), and the glass transition point (Tg) of the polymer as synthesized was found at 92 °C (DSC). Poly(glycidyl methacrylate) (PGMA) 4 was synthesized in bulk at 120 °C by nitroxide-mediated polymerization (NMP) (Scheme 2). A solution of 24.6 mg of alkoxyamine initiator 3 (75 µmol) in 2.5041 g of glycidyl methacrylate (18 mmol; Fluka, purum) was prepared in a Schlenk tube and degassed by the freeze-pump-thaw method. The polymerization was conducted at 120 °C for 17 h working under nitrogen atmosphere. The solidified reaction mixture could be dissolved in a small portion of chloroform (VWR), and the polymer was collected by repeated precipitation in ethanol (VWR). 1H NMR (500 MHz, 25 °C, CDCl , δ ppm): 4.31 and 3.81 (m, 3 H3), 3.23 (s, H4), 2.84 and 2.63 (s, H5), 2.03-1.29 (m, H1), 1.09 and 0.94 (s, H2). GPC (PS-calibration, CHCl3): Mn ) 30 600 g · mol-1, Mw ) 61 100 g · mol-1; PDI ) 2.0. The solvents methyl ethyl ketone (MEK; Merck) and 4-methyl2-pentanone (Fluka) were used without further purification. Substrates. Highly polished silicon wafers (100 orientation) exhibiting a native silicon dioxide layer (1-2 nm thick) were used as substrates. The wafers were cleaned twice in an ultrasonic bath in dichloromethane for 5 min each time. In an additional step, the substrates were placed in a hydroperoxide solution mixture [Millipore water, hydrogen peroxide (30%), and ammonia solution (25%), ratio 1:1:1] for 30 min to produce a large amount of hydroxyl groups on the surface. The wafers were carefully rinsed several times with Millipore water. After that, the water contact angle of these silanolrich “activated” silicon wafers was near 0°. Film Preparation. All films were prepared by spin coating in a clean room at 24 ( 1 °C and a relative humidity (RH) of about 50%. The spin coating was carried out using a Spin 150 single-wafer spinner (SPS Europe B.V.) at a spinning velocity of ω ) 3000 rpm for 30 s for aaPOH and at ω ) 2000 rpm for 20 s for PGMA. The aaPOH was dissolved in 4-methyl-2-pentanone (1 wt %). For the grafting, the PGMA layer was prepared from 0.02 wt % MEK solution. Before coating, the solutions were filtered through a 0.2µm Teflon filter in order to remove any undissolved polymer or dust particles. Electron Beam Irradiation. The electron beam irradiation was performed with the ELV-2 electron accelerator (Budker Institute of Nuclear Physics, Nowosibirsk, Russia).34 The accelerator was operated at a beam energy (W) of 1.0 MeV at a constant beam current of 4 mA. The aaPOH samples were irradiated at 200, 400, and 800 kGy and were treated stepwise (e.g., single dose of 66 kGy, (30) Weberskirch, R.; Hettich, R.; Nuyken, O.; Schmaljohann, D.; Voit, B. Macromol. Chem. Phys. 1999, 200, 863–873. (31) Komber, H.; Ziemer, A.; Voit, B. Macromolecules 2002, 35, 3514–3519. (32) Schmaljohann, D.; Komber, H.; Voit, B. Acta Polym. 1999, 50, 196–204. (33) Lederer, A.; Voigt, D.; Clausnitzer, C.; Voit, B. J. Chromatogr. A 2002, 976, 171–179. (34) Dorschner, H.; Jenschke, W.; Lunkwitz, K. Nucl. Instrum. Methods Phys. Res. B 2000, 161-163, 1154–1158.
9394 Langmuir, Vol. 24, No. 17, 2008
Reichelt et al. Scheme 2. Synthetic Approach to PGMA by NMP
total dose of 200 kGy). Depending on the dose D and dose rate DR, the maximum temperature of the samples varied. At adiabatic conditions, the maximum temperature increase due to electron treatment can be estimated according to eq 1. High dose rates and high doses lead to higher temperatures inside the film.
DR )
dD dW c∆T ) ) dt dmdt ∆mdt
(1)
The immobilization of the aaPOH films was carried out under ambient conditions at room temperature and at 100 °C. The absorbed doses were adjusted between 10 and 800 kGy. After irradiation, the films were extracted with 4-methyl-2-pentanone. Grafting-to Procedure. The grafting of aaPOH requires additional steps in the process.1 We developed a straightforward procedure to produce stable films of variable thickness (Figure 1). Thin spincoated PGMA anchoring layers on silicon substrates (step 1) were deposited onto a heated plate for 5 min (step 2). Nonbound components were removed by washing with MEK (step 3), followed by air drying. Then, the aaPOH layer was applied by spin coating as a second layer on top (step 4). Additionally, thin films of aaPOH were immobilized directly on activated silicon substrates without using the adhesive anchoring layer (“self-adhesion”). To force the grafting-to process of the HBP layer, the samples were annealed on the hot plate again for 10 min (step 5). The annealing was performed in a vacuum oven for an additional 2 h (step 6). The annealing of the anchoring (step 2) and HBP (step 5) layers was carried out at the same temperature. Nongrafted polymer was removed by extraction of the films in 4-methyl-2-pentanone for 2 h (step 7). Finally, the samples were dried for 30 min at 40 °C in a vacuum oven to remove remaining solvent. For the comparison of the immobilization technique and determination of its success, the refractive-index- and thickness-dependent degrees of immobilization DIn and DId, respectively, were determined, as defined in eqs 2 and 3
DIn )
ne n0
(2)
DId )
de d0
(3)
where n and d are the refractive index and thickness of the immobilized films obtained by spectroscopic ellipsometry before (subscript 0) and after (subscript e) extraction. Ellipsometry. The thicknesses of the thin covalently immobilized PGMA anchoring layers were measured at an incident angle of 70° and λ ) 632.8 nm with a SENTECH SE-402 microfocus ellipsometer. Single-wavelength ellipsometry is a very sensitive tool for very thin films, but in this case, it was difficult to determine both the thickness d and the refractive index n from only the two measured ellipsometric
values ∆ (relative phase shift) and tan Ψ (relative amplitude ratio). Thus, n must be known or assumed. Here, n for the PGMA layer was set to 1.525.35 Spectroscopic ellipsometry measurements of the aaPOH films were performed on a multiwavelength J.A. Woollam M-2000 rotating compensator ellipsometer in the wavelength range of 371-1679 nm at incident angles of 65°, 70°, and 75°. To fit the n and d values, a multilayer model was assumed, consisting of silicon, silicon dioxide, and polymer layers. A Cauchy relation describes the dependence of n of the nonabsorbing polymer on the wavelength λ. A thickness map of a microstructured aaPOH film was obtained by imaging ellipsometry using an EP3 imaging ellipsometer (Nanofilm Technologie GmbH, Go¨ttingen, Germany). First, the ellipsometric contrast images were measured by a CCD camera (768 × 572 pixel) at an incident angle of 65° and λ ) 632.8 nm. Applying a convenient optical model from selected “regions of interest”, one can map the entire image, yielding a two-dimensional ∆/Ψ map, which can be transferred (using the refractive index data obtained from spectroscopic ellipsometry) into a thickness map.36All samples were examined in the dry state at 23 ( 1 °C and a RH of 50 ( 5%. X-ray Photoelectron Spectroscopy. The chemical composition of the films was studied by X-ray photoelectron spectroscopy (XPS) using an AXIS Ultra instrument (Kratos Analytical, Manchester, U.K.) equipped with a monochromatic Al KR1,2 X-ray source of 300 W at 15 kV. The kinetic energy of the electrons was analyzed with pass energies of 160 eV for survey and 20 eV for high-resolution spectra. The take-off angle, defined here as the angle between the sample surface normal and the electron-optical axis of the spectrometer, was 0°. The information depth of the XPS method is no more than 10 nm at a maximum.37,38 During all measurements, electrostatic charging of the sample was overcompensated by means of a low-energy electron source working in combination with a magnetic immersion lens. Later, all recorded peaks were shifted by the same amount, which was necessary to set the C 1s peak to 284.70 eV.39 Quantitative elemental compositions were determined from peak areas using experimentally determined sensitivity factors and the spectrometer transmission function. The high-resolution C 1s spectra were deconvoluted into four component (A-D) and shake-up peaks, which were related to the chemical structure of the polyester (cf. Figure 2). The shake-up peaks result from π f π* transitions in the aromatic ring. To determine the amount of accessible hydroxyl groups, a labeling reaction with trifluoroacetic anhydride was performed. The amount of CsOH groups in the upper layer of the aaPOH films can be calculated by the difference of the ratios of the atomic concentrations of components D and C. Surface Wettability. Static advancing and receding water contactangle measurements were carried out using drop-shape analysis (DSA 10, Kru¨ss, Hamburg, Germany). For each determination, at least 12 measurements for averaging the contact angles were carried out.
Figure 1. Protocol of the thermal grafting procedure with and without PGMA adhesive layer, according to ref 1.
Immobilization of a Hyperbranched Polyester
Langmuir, Vol. 24, No. 17, 2008 9395
Figure 2. Deconvolution of a C 1s spectrum recorded for a typical aaPOH film applied on a silicon wafer and assignment of the component peaks A-D to the carbon atoms of aaPOH.
The measurements were done under constant conditions (T ) 23 ( 1 °C, RH ) 50 ( 2%). Atomic Force Microscopy. Atomic force microscopy (AFM) was performed with a Dimension 3100 scanning probe microscope (Digital Instruments, Inc., Santa Barbara, CA) in tapping mode. The spring constant of the tip was about 40 mN/m, and the radius was smaller than 10 nm. Electrokinetic Measurements. Streaming potential measurements were carried out using an EKA electrokinetic analyzer from the Anton Paar KG, Graz, Austria. According to the Smoluchowski equation, the apparent zeta potential (ζapp) was calculated from the streaming potential values measured as a function of the pH of an aqueous KCl test solution (cKCl ) 3 × 10-3 mol · L-1). The zeta potential that exists in the shear plane of the electrochemical double layer near the surfaces is correlated with the net surface charge. The pH value where the zeta potential is zero is called the isoelectric point (pH|ζ ) 0 ) IEP). Dynamic Scanning Calorimetry. We used a DSC Q1000 instrument (TA Instruments, New Castle, PA). The samples were weighed into aluminum pans measured under nitrogen atmosphere in the temperature range from - 60 to 200 °C at a heating rate of (20 K/min. The glass transition point was taken from the second heating cycle. Size-Exclusion Chromatography. The SEC measurements of aaPOH were carried out in tetrahydrofuran (THF) using a Knauer SEC apparatus (Berlin, Germany) coupled with an RI and multiangle lightscattering (MALLS) detection with one PL Mixed C column (300 mm × 7.5 mm).
Figure 3. Influence of the dose [applied stepwise (9) and in one step with 200 kGy (0)] on DId of electron-beam-immobilized aaPOH films (dnonextracted ) 20 nm).
3. Results and Discussion 3.1. Immobilization of Hyperbranched Polyester by Electron Beam Irradiation. The aaPOH films coated from 1 wt % solution yielded a typical thickness of about 20 nm. They were homogeneous and flat (cf. Figure 5 below). The samples were irradiated without further pretreatment. Through the irradiation, radicals were generated predominantly from the methylene groups in the aliphatic backbone part. Two effects cause the immobilization of the polymer film: (1) intermolecular cross-linking initiated by macroradicals and (2) linkage of the polymer to the silicon substrate. The grade of the immobilization by electron treatment is influenced by several experimental parameters such as dose, dose rate, and temperature. An increase of the dose leads to higher cross-linking and improved immobilization, but at higher doses, degradation takes place concurrently. In the case of lower doses, only a few molecules can be bound to the SiOx surface. The major difficulty is finding the dose range that is sufficient (35) Van Krevelen, D. W. Properties of Polymers; Elsevier: Amsterdam, 1997. (36) Schmaljohann, D.; Nitschke, M.; Schulze, R.; Eing, A.; Werner, C.; Eichhorn, K.-J. Langmuir 2005, 21, 2317–2322. (37) Briggs, D. Characterization of Surfaces; Pergamon Press: Oxford, U.K., 1989. (38) Seah, M. P.; Dench, W. A. Surf. Interface Anal. 1979, 1, 2–11. (39) Beamson, G.; Briggs, D. High Resolution XPS of Organic Polymers, The Sienta ESCA 300 Database; J. Wiley & Sons: Chichester, U.K., 1992.
Figure 4. High-resolution C 1s XPS spectra of differently electronbeam-irradiated aaPOH. Dose: (a) 100, (b) 200, and (c) 400 kGy. The origin of component peaks A-E is explained in the text (Figure 2).
for the immobilization of the polymer film and leads to minor polymer degradation. The thin films were irradiated at doses between 10 and 800 kGy. In Figure 3, the resulting DId values of the electron-beam-immobilized films are displayed. The thickness of the nonextracted, irradiated films was about 20 nm. As expected, a strong influence of the dose was observed. The higher the dose, the higher the thickness of the immobilized film after extraction. Because of the increase of the dose (eq 1), more energy is applied into the film, resulting in a higher radical yield, which causes higher cross-linking density. The application of doses below 100 kGy is not sufficient for the formation of new bonds inside the layer. Information about the chemical composition of the irradiated films were taken from the high-resolution XPS spectra of the C 1s peak (Figure 4) and the survey spectra. The results revealed that, at higher doses (D g 400 kGy), oxidation
9396 Langmuir, Vol. 24, No. 17, 2008
Reichelt et al.
Figure 5. AFM height images of aaPOH films: (a) nonirradiated (scale, 5 nm), extracted; (b) irradiated at D ) 100 kGy (scale, 15 nm); (c) irradiated at D ) 200 kGy (scale: 5 nm). Table 1. Quantitative Evaluation of XPS Data Showing Elemental and Component Peak Ratio of Untreated and Electron-Beam-Treated aaPOH
Table 2. Contact Angles, IEPs, Mean Refractive Indexes, And Layer Thicknesses of Nonirradiated and Irradiated aaPOH Films θa (deg)
electron-beam treated untreated [O]/[C] D/C
0.221 0.499
100
kGy
0.331 0.506
200
kGy
0.346 0.467
processes led to the formation of carboxylic acid groups. The beginning of polymer degradation was proven by an additional component peak E (appearing from newly created carboxylic acid groups) in the C 1s peak and by a significant decrease of the hydrophobicity (DSA). Compared to the C 1s peak of the XPS spectra of nonirradiated polymer films, the intensities of the shakeup peaks were clearly lower, caused by destruction of conjugated π-orbitals. Therefore, for all further irradiation experiments, only limited doses of 100 and 200 kGy were chosen. The derived ratios of component peaks D/C (cf. Figure 2) and the elemental ratio [O]/[C] of untreated and 100- and 200-kGy-irradiated polymers are compared in Table 1. The component peak area ratios of D/C ) 1:1.98 and 1:2.14, respectively, match the theoretical value of 0.5 quite well (cf. Figure 2).1 The oxygen content of the polymer film increases as a result of the reaction with oxygen radicals (cf. Table 1). Nevertheless, the thickness and refractive index of the nonirradiated and irradiated, nonextracted samples were nearly identical, clearly indicating that no gaseous degradation products were formed that normally would lead to film shrinking (thickness decrease) and/or pore formation (refractive index decrease). As for the nonirradiated polyester films, we could also confirm for the irradiated and extracted films uniform surface covering with roughness in the range of the substrate roughness (below 10 nm) on PGMA should show a combined effect of surface grafting to PGMA and the self-adhesion mechanism (step II) described next. (ii) Self-Adhesion. Here, films of aaPOH were spin-coated on activated silicon surfaces without additional use of any anchoring layer. This one-layer system was annealed in a first step on the hot plate for 10 min and in the second step in a vacuum oven at elevated temperatures above 140 °C. After immersion in 4-methyl-2-pentanone, the polymer films surprisingly remained on the substrate. The optimal immobilization temperature was found to be 160 °C, similarly to the grafting to PGMA. An increase of the annealing time (Figure 8A) results in an increase in the amount of immobilized aaPOH. The samples were annealed additional 2, 4, 24, 30, and 72 h in vacuum oven. The total thickness values are shown in Figure 8B. Similarly to the PGMA grafted polyesters, in this case also, quite thick films could be immobilized after 8 h of treatment at 160 °C, although, to obtain the same quality of immobilized films, longer annealing times were necessary (cf. Figure 9). Homogeneous and smooth film formation was confirmed by AFM topography measurements (Figure 10). Information on the chemical composition were obtained from XPS spectra recorded on the trifluoroacetic anhydride- (TFAA)labeled aaPOH films immobilized by self-adhesion (Figure 11). The calculated [F]/[C] ratios, which are a verification of free OH groups, were constant until an annealing time of 24 h. The shape of the C 1s spectra did not show any significant changes. Further thermal treatment resulted in a significant decrease of the [F]/[C] ratio from 0.105 to 0.74 and lower intensities for the component peaks appearing from the trifluoroethyl ester group. We also succeeded in the preparation of 350-nm-thick immobilized aaPOH layers. The only limitation on the preparation
Figure 9. Dependence of the thickness of aaPOH films immobilized by self-adhesion (annealing for 8 h at 160 °C) on the concentration of the polymer solution used for spin-coating.
Figure 10. AFM topography image of a 20-nm-thick aaPOH film immobilized by self-adhesion (annealing for 72 h at 160 °C).
of thicker films seems to be (1) the increasing inhomogeneity and roughness of the layers with increasing thickness and (2) the limited solubility of aaPOH in the solvent 4-methyl-2-pentanone chosen for spin-coating at high concentrations. The self-adhesion proceeds in two steps. In the first step, the polymer is anchored to the Si substrate; the second step is the stabilization inside the polymer film. (1) Anchoring to the Silicon Substrate. One possible explanation is the formation of a strong hydrogen-bond network. Hydrogen bonds between the silanol groups of the substrate and another polar group (e.g., phenolic group, ester group) are widely
Immobilization of a Hyperbranched Polyester
Langmuir, Vol. 24, No. 17, 2008 9399 Table 5. Characterization of the Effects of Annealing on Bulk Samples of aaPOH
Figure 11. High-resolution XPS C 1s spectra of thermally cured aaPOH films labeled with TFAA: (a) thermally untreated aaPOH film and aaPOH films annealed at 160 °C for (b) 2, (c) 4, (d) 24, and (e) 72 h. All component peak intensities were referred to the area of component peak A. The unmarked component peaks are shakeup peaks.
believed to influence polymer-solid adhesion.43 Keddie et al. stated that, in the interaction between poly(methyl methacrylate) and the native oxide layer of a silicon substrate, hydrogen bonds play a significant role. The restriction of the chain mobility leads to an increase of the glass transition temperature.44Sidorenko et al. investigated the self-assembly of aliphatic hyperbranched polyesters from acetone solution.13 They stated that, after a simple washing step, bilayers of the polymers (thickness of approximately 3 nm) were formed by weak hydrogen bonds. Our investigations showed that new hydrogen bonds were formed in an aromatic HBP at elevated temperatures and that they widely influenced the film properties.45The strength of the hydrogen bond lies in the range of 20 kJ mol-1 (cf. 331 kJ mol-1 for a CsO bond).46 However, hydrogen bonds cannot be the only explanation for fixing the polymer at the substrate. Already in 1961, Ballard et al. investigated the mechanism of esterification of alcohols with surface silanols of amorphous silica.47 In our system, large amounts of both aaPOH hydroxyl groups and surface silanol groups are available for reaction. Also, we can not exclude the possibility that reaction between the few carboxylic groups of the polyester (coming from the valeric acid monomer) and the silanol takes place. Unfortunately, the proposed chemical reaction (43) Gutowski, W. In Fundamentals in Adhesion; Lee, L.-H., Ed.; Plenum Press: New York, 1991; p 87. (44) Keddie, J. L.; Jones, R. A. L.; Cory, R. A. Faraday Discuss. 1994, 98, 219–230. (45) Mikhailova, Y.; Adam, G.; Ha¨uβler, L.; Eichhorn, K.-J.; Voit, B. J. Mol. Struct. 2006, 788, 80–88. (46) Carey, F. A.; Sundberg, R. J. Organische Chemiesein weiterfu¨hrendes Lehrbuch; Wiley-VCH: Weinheim, Germany, 1995. (47) Ballard, C. C.; Broge, E. C.; Iler, R. K.; St. John, D. S.; McWhorter, J. R. J. Phys. Chem. 1961, 65, 20–25.
aaPOH
θa (deg)
IEP
Tg (°C)
untreated 5h 8h 24 h 72 h
78.1 ( 0.2 80.6 ( 0.7 80.8 ( 0.1 83.3 ( 0.7 88.8 ( 1.5
3.8 4.0 3.9 4.2 4.0
92 129 130 131 140
of the surface silanol groups with the hydroxyl groups of the aaPOH layer cannot be followed directly, such as by spectroscopic methods. Therefore, in a model experiment, the formation of stable films was suppressed by sputtering a thin but optically dense gold coating on the silicon dioxide layer. As a result, any interaction between the surface silanol groups and the polymer was inhibited. The aaPOH was spin-coated on top and treated thermally as before, but in this case, extraction in the solvent led to the total removal of the polymer film from the gold surface. Probably, all three discussed interactions occur simultaneously and contribute to the formation of anchored polymer films, but the experiment on the gold surface allows for the conclusion that the silanol groups of the silicon substrate surface are an essential condition for the immobilization of aaPOH without a PGMA anchoring layer. (2) Stabilization inside the Polymer Layer. We claim that the stabilization of the film is caused initially by hydrogen bonds and that, after longer annealing times, transesterification and ether formation between available phenolic OH groups of the repeating units will take place. Malmstro¨m et al. stated, during an investigation of 2,2-(bishydroxymethyl)propionic acid based hyperbranched polyesters, that ether formation under acid conditions is possible.48 However, intermolecular transesterifications catalyzed by the acidic phenolic hydroxyl groups should play a more important role. Some ester bonds were cleaved, and others were formed, so the total amount of ester bonds remains nearly constant. The fact that, apart from a decrease of the OH group content (at high annealing times), there are no other marked changes in the XPS spectra of immobilized (annealed at 160 °C for different times) aaPOH layers compared to the nonimmobilized (nonannealed) layers supports this assumption. Both transesterification and ether formation lead to insoluble network structures within the aaPOH layers. Our explanation of the mechanism of self-adhesion of aaPOH layers is supported by results obtained for bulk samples (powder) that were prepared by annealing in a vacuum dryer at 160 °C for 2, 4, 24, and 72 h. The annealing led to a shift of the glass transition temperature (determined by DSC) to higher values especially for longer annealing times (Table 5). It is known that the formation of a physical or chemical network leads to an increase in Tg.44 This is caused by the restriction of the mobility of the polymer molecules, which seems to have an effect also on some of the surface properties. Therefore, we actually found an increase of the contact angle for annealing times above 24 h because of the consumption of surface hydroxyl groups; hence, the IEP did not change significantly (Table 5). A further hint for a cross-linking reaction was the decrease in solubility in common organic solvents with increasing annealing time. As shown in Figure 12, the molar mass and polydispersity of the soluble part of annealed bulk samples was lower than of the untreated bulk aaPOH. However, the typical shape of the refractive index (RI) curve remained unchanged. (48) Malmstro¨m, E.; Hult, A. Macromolecules 1996, 29, 1222–1228.
9400 Langmuir, Vol. 24, No. 17, 2008
Figure 12. Refractive index (RI) and light scattering (LS) signals of the GPC measurements on bulk aaPOH samples: untreated (s) and annealed for 5 h (- -) and 8 h (- · -) (T ) 160 °C).
IR spectroscopic analysis of a 72-h-annealed bulk sample showed only slightly changes in the carbonyl band profile of the ester groups. This points to a decrease of weakly associated hydrogen bonds,45 but also to cleavage and re-formation of some ester bonds themselves due to transesterification reactions. In summary, we assume that the structure of aaPOH in the bulk state in the first 24 h of annealing is strongly influenced by the formation of a hydrogen-bond network. Longer annealing times lead to chemical network formation through transesterification, ether formation, and further condensation.
Conclusion We elaborated two novel techniques of immobilization of an aromatic-aliphatic hyperbranched polyester aaPOH on silicon substrates in such a way that the surface properties of the thin films and the high functionality of the polymer remain more or less unchanged in comparison to those of nongrafted polyester films produced by simple spin coating. First, thin films of aaPOH were effectively immobilized and partially cross-linked by electron beam treatment on silicon
Reichelt et al.
wafers. The method was optimized to obtain optimal nondegenerative immobilization conditions. In principle, different experimental conditions (dose, dose rate, temperature) can influence the quality of immobilization. Our results show that especially the temperature widely influences the main process. Doses higher than 200 kGy result in degradation and oxidation. The use of the method for the easy preparation of structured surfaces was demonstrated. The advantage of this method is that the immobilization occurs within seconds to a few minutes without the use of additional coupling agents or thermal treatment. Another fast and effective method of aaPOH immobilization is a heat-stimulated multistep grafting process on a thin intermediate layer of poly(glycidyl methacrylate) (PGMA), which is a well-known polymeric coupling agent for the formation of polymer brushes through the grafting-to process of linear carboxyl-terminated polymers. Thin films of aaPOH could also be strongly bound without a coupling agent to an activated silicon surface by a simple annealing step on a hot plate and additional annealing in a vacuum oven. In comparison to the immobilization of aaPOH on PGMA layers, longer grafting times are necessary to obtain the same degrees of immobilization. We suppose that the direct anchoring to the silanol groups of the silicon dioxide layer is the determining step for this kind of immobilization, in addition to the formation of an insoluble network within the polymer layer. The novel method was optimized for conditions such that no degradation of the polymers occurred. This method might also be of interest for the preparation of stable thicker coatings, as no additional chemistry is necessary. The preparation of thicker immobilized films on PGMA and by self-adhesion is caused by physical and chemical cross-linking. Acknowledgment. The authors are pleased to thank Mrs. L. Ha¨uβler and Mrs. K. Arnhold for the DSC measurements, Mrs. S. Dziolloss for GPC measurements, Dr. C. Bellmann and Mrs. A. Caspari for electrokinetic measurements, and Mr. S. Polnick for supporting the preparation of the films. This work was financed by the Deutsche Forschungsgemeinschaft in the framework of SFB 287 “Reactive Polymers”. Dr. K. Grundke is acknowledged for help with the interpretation of contact-angle data. LA8006662
