Polymers tethered to substrates represent a central part of a wide range of applications, including stabilization of colloids, adhesion promotes, and polymer coatings. In addition these systems are also attractive from the fundamental point of view – they provide unique means of probing the interplay between the polymer thermodynamics and the confinement effect due to the attachment to the substrate.
The topic of neutral tethered polymer chains has received considerable attention – both theoretical and experimenta. The major findings of the various theoretical and experimental approaches have revealed that depending on the grafting density of the polymers at the solution/substrate interface, σ, the anchored chains form either so-called mushrooms or so-called brushes. In good solvents, the thickness of the anchored polymer, H, in the low grafting density mushroom regime scales as H~N0, where N is the degree of polymerization of the polymer; in the brush regime the chains become more crowded and the brush height scales as H~N1/3.
Multiple research reports appeared that attempted to test the above scaling relations and the conditions for the mushroom-to-brush transition. The typical hurdles most experimental studies usually run into involve complications associated with: 1) preparing brushes with high enough grafting densities, and 2) exploring completely the broad σ space. The major objective of our work was to develop a method that would circumvent these two obstacles simultaneously. Specifically, we demonstrated that by preparing density gradients of polymerization initiators on substrates and using “grafting from” polymerization on such surfaces, arrays of anchored polymers with a gradual variation of grafting densities can be fabricated.
In order to prepare anchored polymers with gradually varying grafting densities on solid substrates, a gradient of 1-trichlorosilyl-2-(m/p-chloromethylphenyl) ethane (CMPE), the polymerization initiator, was first formed on the surface using the vapor deposition technique, proposed by Chaudhury and Whitesides [Science 256, 1539 (1992)] (paraffin oil, PO, was used as a diluent for CMPE).
In order to minimize any physisorption of monomer and/or the polymer formed in solution on the parts of the substrate that do not contain the CMPE-SAM, the unexposed regions on the substrate containing unreacted –OH functionalities were treated with n-octyl trichlorosilane (OTS). Near-edge x-ray absorption fine structure (NEXAFS) spectroscopy was utilized to measure the concentration of the CMPE along the CMPE-SAM gradient.
The concentration of CMPE in the sample decreased as one moved from the CMPE side of the sample towards the OTS-SAM; the functional form closely resembled that of a diffusion-like profile (solid line in the figure on the right). Experiments using variable angle spectroscopic ellipsometry (VASE) confirmed that only a single SAM monolayer was formed on the substrate. Monodisperse poly(acryl amide) (PAAm) chains with a gradual variation in grafting densities were synthesized by “grafting from” reaction of acryl amide using the atom transfer radical polymerization (ATRP). VASE was utilized to measure the thickness of the dry polymer film, h, as a function of the position on the substrate. Because the polymers grafted on the substrate have all roughly the same degree of polymerization, the variation of the polymer film thickness can be attributed to the difference in the density of the CMPE grafting points on the substrate.
The data in the figure above reveal that h decreases gradually as one moves across the substrate starting at the CMPE edge. Note that the concentration profile of the polymer follows that of the initiator.
The substrates with the grafted PAAm were placed into a solution cell that was filled with deionized (DI) water (pH~7), a good solvent for PAAm, and incubated for at least 5 hours. The wet thickness of PAAm grafted polymer in DI water, H, was measured using VASE. The values of H for samples prepared on CMPE:PO=1:1 gradients are shown in the figure above. The data illustrate that H decreases as one traverses across the substrate starting at the CMPE side. In the figure on the right we plot the wet polymer thickness as a function of the PAAm grafting density on the substrate. At low σ, H is independent of the grafting density. Hence the chains are in the mushroom regime. At high polymer grafting densities, H increases with increasing σ, indicating the brush behavior.
The cross-over between the two regimes occurred at σ~0.065 nm-2. By fitting the data in the brush regime to H~Nn we get n equal to 0.37±0.04 (CMPE:PO=1:1), 0.39±0.05 (CMPE:PO=1:2), and 0.40±0.06 (CMPE:PO=1:5); in close agreement with the predicted value of n=1/3. A remark has to be made about the possible variation of the chain length with grafting density. While we cannot exclude the possibility that the length PAAm chains polymerized on the various parts of the molecular gradient substrate varies with σ, we note that the fact that the curves in the figure on the left superimpose on a single master curve indicates that the polymers have likely very similar lengths, which is not surprising for the rather short anchored polymers synthesized in this work.
In addition to the measurement of the wet brush thickness, we have also performed wettability experiments (by measuring the contact angle of the DI water θDIW), as a function of the PAAm grafting density on the substrate. Our aim was to corroborate the ellipsometric data and provide more insight into the polymer packing in the surface grafting density gradient.
In the figure on right we plot the negative cosine of θDIW as a function of the grafting density of PAAm on CMPE:PO=1:1 and CMPE:PO=1:5 substrates. As anticipated, the data collapse on a single master curve. A close inspection of the results shows that the data can be divided into three distinct regions. For σ>0.1 nm-2, the chains are expected to be in a brush regime – the wettabilities are close to the pure PAAm (-cos(θDIW)~-0.79). For σ<0.011 nm-2 the PAAm chains form mushroom conformations on the substrate. In this regime, the wettabilities change slightly because the distance between the chains also changes, although they are already loosely separated on the substrate. At grafting densities 0.011 nm-2<σ<0.1 nm-2, the slope of -cos(θDIW) changes rather rapidly.
The data show that the position of the mushroom-to-brush crossover determined using the wettability approach is in accord with the ellipsometric measurements (located by the arrow in the figure). However, in the former case, the transition region extends over almost one order of magnitude in σ, which is broader, as expected, that the transition region predicted by the H vs. σ data. We speculate that the small difference between the widths of the mushroom-to-brush region inferred from both types of experiments is likely associated with the inaccuracy in H, which was obtained indirectly by the model fitting of the VASE data.