It is well known that when polymers are end-anchored in a sufficient concentration to a substrate, they form a so-called polymer “brush” whereby the chains are stretched with respect to their preferred configuration away from the interface. The characteristics of polymer brushes have been analyzed using a variety of theoretical methods and experimental probes and are now fairly well established. For example, the thickness of the brush layer, H, is known to depend linearly on the number of repeat units of the polymer, N, and on the power law of the brush grafting density at the substrate, σ, with the exponent of 1/3 or 1, depending on the surface coverage. In order to fine-tune the polymer brush properties one needs to have a good control over H and σ. While H can be adjusted by simply varying the polymerization time, monomer concentration and monomer conversion, σ depends on the methods by which polymer brushes are formed.
Previous reports established that in contrast to “grafting onto” methodologies, polymer brushes with moderately high grafting densities can be prepared by harnessing the “grafting from” principle in which the polymer chains are synthesized using radical initiators that are covalently bound to the substrate. A vast majority of experiments involved classical radical growth methods with either azo- (e.g., AIBN) or peroxide-based initiators that were either created directly on the substrate or were attached to the substrate via self-assembly. Recently several reports appeared that described the formation of polymer brushes using surface-initiated “living” radical polymerization, such as the atom transfer radical polymerization (ATRP). Because of its simplicity, robustness and the ability to synthesize polymers with narrow molecular weight distributions, ATRP has been the method of choice for most surface-initiated “living” radical polymerization processes.
One of the crucial parameters governing the behavior of polymer brushes is their grafting density at the polymer/substrate interface. The “grafting from” techniques offer a fairly good control over the grafting density of the polymer brush – in the ideal case σ is simply equal to the surface density of the polymerization initiators. While a relatively high density of initiators can be achieved by assembling the molecules on the surface by means of Langmuir-Blodgett (LB) techniques or by forming organized self-assembled monolayers (SAMs), tailoring the grafting density of the SAM chains is not an easy task. SAMs are usually formed through self-assembly processes that are governed by the chemical and structural nature of the SAM molecules and the means of their attachment to the substrate. To overcome this limitation, one would need to seek another way of controlling the grafting density of the surface initiators that is independent of the system thermodynamics.
We have recently shown that polymer brushes with high grafting densities can be formed by utilizing mechanically assisted polymer assembly (MAPA) principle. The figure above shows schematically the preparation of poly(acryl amide) brushes using MAPA. First, a pristine PDMS network film is prepared by casting a mixture of PDMS and a cross-linker into a thin (ca. 1 mm) film and curing it at 70°C for about an hour. The film is then cut into small strips (ca. 1×5 cm2) and mechanically elongated by Δx. subsequent exposure to UV/ozone (UVO) treatment produces hydrophilic PDMS surfaces (PDMS-UVO) composed mainly of hydroxyl groups (HO-[Si]surface), that serve as attachment points for chlorosilane-based ATRP initiators. Following previous work on PAAm brushes, we used 1-trichlorosilyl-2-(m-p-chloromethylphenyl)ethane (CMPE) as the initiator. The CMPE molecules were chemisorbed from vapor onto this stretched substrate. The PAAm ATRP on the PDMS-UVO/CMPE substrates is performed as described previously. [Huang et al. Chemtech Dec 1998, 19].
We used near-edge absorption fine structure (NEXAFS) spectroscopy to examine the surface and bulk chemistry (including bond densities) of the samples. The NEXAFS experiments were conducted at the at the Soft x-ray Materials characterization facility at the National Synchrotron Light Source at Brookhaven National Laboratory. By simultaneously collecting both the partial electron yield (PEY) and fluorescence (FY) NEXAFS signals, whose probing depths are ca. 2 and ca. 100 nm, respectively, the surface and bulk chemical compositions of the sample can be probed in a single experiment. Because no measurable nitrogen signal could be detected in the nitrogen K-edge FY NEXAFS spectra and also because the carbon K-edge and oxygen K-edge FY NEXAFS spectra were almost indistinguishable from that of bare PDMS, the NEXAFS measurements confirmed that the PAAm brushes were present only on the sample surface and not in the bulk.
The figure on the right shows the PEY NEXAFS spectra taken at the nitrogen K-edge (left) and oxygen K-edge (right) of PDMS-UVO and PAAm-MAPA samples. The absence of any nitrogen signal for PDMS-UVO and strong absorption peaks detected at the nitrogen K-edge of the PAAm-MAPA samples verify that PAAm brushes were formed during the MAPA process. This finding is further supported by exploring the PEY NEXAFS spectra collected at the oxygen K-edge. While the broad peaks in the oxygen K-edge between 534–548 eV that are present in all spectra can be associated with various bonding environments of Si-O, the peak at 531 eV corresponding to the 1s-&grt;π*C=O transition is detected only in the PAAm-MAPA samples, providing an additional evidence for the presence of PAAm.
A close inspection of the data reveals that the intensity of the 1s-&grt;π*C=O transition increases with increasing Δx indicating that the density of the C=O bonds and thus the PAAm chains increases with increasing Δx, as expected.
Further evidence for the observation elucidated from the PEY NEXAFS data was obtained from complementary experiments using Fourier transform infrared spectrometry in the attenuated total reflection (ATR-FTIR). The figure on the left shows ATR-FTIR spectra measured on (top to bottom) PDMS-UVO/CMPE and MAPA-PAAm samples with Δx equal to 0%, 10% and 20%. The spectra were normalized to the Si-CH3 deformation mode at 1414 cm-1. Inspection of the data reveals that there are several characteristic peaks in the ATR-FTIR data. Following previous studies on PAAm, we attribute the 1668 and 1616 cm-1 signals to the symmetric amide C=O stretching mode, and the combination of the C-N stretching/N-H deformation modes, respectively. Moreover, the bands at 3335 and 3200 cm-1 are assigned to the asymmetric and symmetric N-H stretching modes of the amide, respectively.
As apparent from the figure, the intensity of both the C=O and N-H stretching modes increases with increasing Δx. This result thus demonstrates unambiguously that the polyacrylamide grafting density at the PDMS-UVO substrate increases with increasing elongation on the untreated elastic substrate.