Sensing surface PEGylation with microcantilevers

Abstract

Polymers are often used to modify surface properties to control interfacial processes. Their sensitivity to solvent conditions and ability to undergo conformational transitions makes polymers attractive in tailoring surface properties with specific functionalities leading to applications in diverse areas ranging from tribology to colloidal stability and medicine. A key example is polyethylene glycol (PEG), which is widely used as a protein-resistant coating given its low toxicity and biocompatibility. We report here a microcantilever-based sensor for the in situ characterization of PEG monolayer formation on Au using the "grafting to" approach. Moreover, we demonstrate how microcantilevers can be used to monitor conformational changes in the grafted PEG layer in different solvent conditions. This is supported by atomic force microscope (AFM) images and force-distance curve measurements of the microcantilever chip surface, which show that the grafted PEG undergoes a reversible collapse when switching between good and poor solvent conditions, respectively.

Keywords: AFM; cantilever sensor; polyethylene glycol; polymer brush; reversible collapse; static mode.

Figure 1

(A) Scanning electron microscope image of a silicon microcantilever array consisting of eight cantilevers and two sidebars. (B) Schematic drawing of the sensor instrument: 1 - the measurement cell with a mounted microcantilever array, 2 – optical-read out system comprising vertical cavity surface emitting lasers (VCSELs) and a position sensitive detector (PSD), 3 – data processing and acquisition, 4 – valve selector connected to liquid samples.

Figure 2

Grafting of mPEG–SH chains on a Au-coated microcantilever surface. (A) Deflection of the sensing (red) and the reference microcantilevers (blue) upon injection of different mPEG–SH dilutions. The vertical dotted lines indicate the beginning of the injection of a new sample concentration. Each curve represents the aligned and averaged data from at least three microcantilevers. (B) Differential deflection of sensing microcantilevers after subtraction of the reference.

Figure 3

Concentration-dependent grafting of mPEG–SH on Au: overlay view of five binding curves. The grey area indicates the period of injection of mPEG–SH solution. The binding curves were obtained at mPEG–SH concentrations of 0.5 μM (1), 10 μM (2), 50 μM (3), 100 μM (4) and 500 μM (5). Each curve was obtained by subtracting the response of the reference cantilevers from at least two sensing cantilevers from within the same array. The resulting five curves were further normalized with respect to the mechanical properties of the cantilevers used (see Material and Methods section). The dashed line in the curve (5) is the extrapolated saturation signal.

Figure 4

(A) Adsorption isotherm of mPEG–SH on Au. Each point (circle) corresponds to the maximum differential signal observed at the following mPEG–SH concentrations: 0.5, 10, 50, 100 and 500 μM. The solid line represents the Langmuir isotherm fitting curve (R² = 0.982). (B) Representative plots of microcantilever deflection versus time. Kinetic curves (black) were fitted using the exponential Langmuir rate law (red). The curves were fitted for the first 30 min of the binding event and then extrapolated for a longer period of time.

Figure 5

Representative force curves obtained by approaching the AFM tip to the Au surface grafted with 20 kDa mPEG–SH. The exponentially decaying long-range repulsive force in the upper curve indicates that the PEG chains are “brush-like” (black squares) in PBS. Inset: The corresponding fit using eq 3 (red line) gives fitted values of 27.4 ± 0.2 nm and 18.1 ± 0.1 nm for L and s, respectively. The data collected beyond a certain D is scattered being less than the minimum detectable force, which is given by the thermal noise of the AFM-cantilever: F_min = (k_BT × k_AFM)^1/2 ≈ 5 pN. The weak repulsion at ~5 nm in the lower curve indicates that the PEG chains are no longer brush-like in 20% 2-propanol and have collapsed to form a layer that is ~5 nm thick (gray squares). The 0.1 nN offset distinguishes the two force curves as being from separate measurements.

Figure 6

AFM images of Au-tethered PEG-layers obtained at different forces in good solvent (PBS; upper row) and in poor solvent (20% 2-propanol in PBS; lower row) conditions. Images in PBS show an improved resolution of the underlying Au surface (on the same area) as the force set point is increased from 30 → 60 → 80 pN indicating penetration/splaying of the PEG layer by the tip. The Au surface is covered again after the force set point is reduced back to 30 pN. The lower series of images were acquired in 20% 2-propanol and do not show any dependence on the force applied. This effect was similar over different areas on the sample (the 30 pN area is different from the 60 and 120 pN area), and implies that the tethered chains have collapsed under poor solvent conditions. All images were obtained on the same microcantilever array chip. The scale bar is 100 nm.

Figure 7

Collapse and restretching of the tethered mPEG–SH-layer observed when switching from good (PBS) to poor solvent conditions, particularly in 20% (A) and 10% (B) 2-propanol in PBS. The duration of each poor solvent injection is shaded in gray. (A) As an in situ reference two microcantilevers in the array were blocked with EG₄–C₁₁–SH. (B) As an external reference, a non-functionalized Au-coated microcantilever array was used. The averaged deflection of eight reference cantilevers was subsequently subtracted from the average deflection of eight sensing cantilevers to obtain the differential response.

Figure 8

Mushrooms versus Polymer Brushes: a sketch illustrating how surface-tethered polymer chains can take on either “mushroom” -like or “brush”-like molecular conformations, depending on how closely packed the polymer chains are. The mushroom regime occurs when the distance between neighboring chains s, is greater than twice the radius of the polymer. The brush regime is encountered when s < 2r and the polymer chains are extended away from the surface at a height of L. The shaded areas emphasize the stochastic nature of the polymer chains where each chain has a high probability to occupy all positions within a given volume.