Article pubs.acs.org/Langmuir
Quantitative Analysis of Polymer Brush Formation Kinetics Using Quartz Crystal Microbalance: Viscoelasticity of Polymer Brush Hirokazu Tanoue,† Norifumi L. Yamada,‡ Kohzo Ito,† and Hideaki Yokoyama*,† †
Graduate School of Frontier Sciences, The University of Tokyo, Chiba 277-8561, Japan High Energy Accelerator Research Organization, Ibaraki 319-1108, Japan
‡
S Supporting Information *
ABSTRACT: Polymer brush formation kinetics was measured by quartz crystal microbalance (QCM). In the QCM measurement anomalous complex frequency shift as a function of graft density was observed: dissipation rate shift ΔΓ first increased, showed a peak, and then decreased despite the graft density of the polymer brush was increasing monotonically. We calculated the shear modulus of the brush layer from the measured complex frequency shift and revealed that the peak of ΔΓ described the crossover from a viscous brush layer to elastic brush layer. The crossover for the poly(ethylene glycol) with a molecular weight of 2000 occurs at around the characteristic graft density of 0.17 chains/nm2 which was revealed from the structure analysis by surface plasmon resonance (SPR) and neutron reflectivity (NR).
■
INTRODUCTION The grafted polymer chains on solid surfaces are called polymer brushes.1 Attention has been paid on polymer brushes as a tool for surface modification, since they provide the characteristic properties such as preventing adhesion on solid surfaces,2−5 enhancing dispersion of nanoparticles,6−8 and lubrication.9−11 One of well-known methods to fabricate polymer brushes is the “grafting-to” method.12 In this method, end-functionalized polymers are grafted onto solid surfaces by physical or chemical adsorption. The static structure of polymer brushes can be measured and analyzed quantitatively by the interferometric analytical methods for interface such as neutron reflectivity (NR)13−15 and ellipsometry.16,17 These interferometric analytical methods can measure very thin brush structures in liquid; however, it is difficult to measure the growth of polymer brush, which often occurs very fast, due to their limited time resolution. In contrast, quartz crystal microbalance (QCM) is able to detect very small structure change at interface with high time resolution. QCM is not only an acoustic wave sensor which detects small mass change18,19 but also an acoustic viscometer to monitor the rheological properties near the interface20−22 when used in liquid. QCM has been used to detect a small mass change in the field of biointerface23−25 but rarely used to measure the structural changes of polymer brushes26,27 in liquid. For the studies of polymer brush, QCM does not only measure the mass change of polymer brush but also detect the internal relaxation of brush swollen with fluids. Thus, QCM has a potential to measure the kinetics of polymer brush formation with high sensitivity but has also difficulty to be used as a quantitative tool for the study of the polymer brush due to the complexity of QCM response. However, QCM can potentially detect, through the mass and viscosity on the sensor © 2017 American Chemical Society
surface, both the changes of graft density and conformation of brush chains, which occur simultaneously as polymer brush grows.28,29 If it is possible to relate the change of brush structure with the complex frequency shift of QCM quantitatively, QCM can be a useful tool to study polymer brushes with high sensitivity and high time resolution. In this study we measured the kinetics of polymer brush formation by the “grafting-to” method using surface plasmon resonance (SPR) and NR in addition to QCM. We analyzed the structure of brushes by SPR and NR and related the structure change to the complex frequency shift measured by QCM quantitatively. We chose a slowly growing polymer brush by the “grafting-to” method so that the growth can also be followed by SPR and NR. The structures determined by SPR and NR are used to understand the unique complex frequency shift of QCM by brush growth.
■
EXPERIMENTAL SECTION
Materials. Poly(ethylene glycol) methyl ether thiol (Mn = 2000 and PDI < 1.1) (PEG-SH) was purchased from Aldrich and used as received. Deuterium oxide (D2O, 99.9%) was purchased from Aldrich. PEG-SH solutions were prepared by mixing PEG-SH with ultrapure water having a resistivity of 18.2 MΩ cm for SPR and QCM measurements and with D2O for NR measurement. PEG brushes were fabricated by grafting thiol-terminated PEG-SH onto gold-covered quartz, LaFSN9 glass, and 1 cm thick quartz30 for QCM, SPR, and NR, respectively. Surface Plasmon Resonance (SPR). We used SPR to monitor the amount of PEG on the gold surface. A 632.8 nm He−Ne laser was
Received: March 21, 2017 Revised: April 19, 2017 Published: April 20, 2017 5166
DOI: 10.1021/acs.langmuir.7b00961 Langmuir 2017, 33, 5166−5172
Article
Langmuir
rate shift ΔΓ were measured over the fundamental and third, fifth, and seventh overtones.
used as light source. We used LaFSN9 high refractive index glass as a substrate, the surface of which was coated with gold (thickness ≈47 nm). In order to promote adhesion, chromium was evaporated (thickness ≈1 nm) between gold and the glass. The opposite side of the glass was attached to a triangular prism via matching oil. SPR was operated in total internal reflection mode using the Kretschmann configuration.31 Reflectivity at a constant incident angle of 56°, which is in the range where the reflectivity changes linearly with the change of resonance angle, was measured and converted to resonance angle change, Δθ. First, the gold surface of SPR substrate was in contact with pure water until the reflectivity became stable. Then the water was replaced with the 250 nM PEG-SH aqueous solution. The experiment was conducted at room temperature as a function of time after injection of PEG-SH solution. The graft density σ [chains/nm2] was calculated by the equation
■
RESULTS The graft density of PEG brush was measured by SPR (Figure 1) as a function of time. The graft density monotonically
σ = 1.68Δθ where Δθ is the shift of the resonance angle in degrees. The coefficient was calibrated to be valid in the range of brush density from 0 to 0.60 [chains/nm2]. Δθ was estimated from the reflectivity change and the slope of the resonance curve.32 The calibration for the above equation is provided in the Supporting Information. Neutron Reflectivity. Neutron reflectometry experiments were conducted using Soft Interface Analyzer (SOFIA)33,34 at Beamline 16 of J-PARC/MLF. We used 1 cm thick quartz wafer with a diameter of 3 in. as a substrate. We evaporated gold on the surface of quartz (thickness ≈10 nm) after chromium was evaporated as adhesive layer (thickness ≈3 nm). We conducted time-resolved measurement for 7200 s after the 250 nM PEG-SH solution in D2O was introduced into a cell using a liquid injection system,35 in which wavelength band from 0.2 to 1.78 nm was used with the incident angle of 1.10°. After 7200 s, reflectivity of equilibrium brush structure was measured at two different incident angles (0.60°, 1.10°): a wavelength band from 0.2 to 1.78 nm was used. We used the data measured at incident angle of 0.60° to convert the data of time-resolved measurement to absolute reflectivity. Finally, we measured reflectivity, in which wavelength band from 0.2 to 0.88 nm was used with the incident angle of 0.30°, 0.75°, and 1.80°, after 20 h from the injection of PEG-SH solution. The scattering length density (SLD) profile normal to the quartz surface was computed by fitting the reflectivity spectra using the MOTOFIT program.36 We fitted the reflectivity spectra with a multilayer model consisting of quartz, chromium, gold, PEG brush layer, and PEG-SH solution. The concentration of the PEG-SH solution is low enough to consider the SLD of the solution is the same as that of D2O. The SLD values of quartz, chromium, gold, PEG, and D2O were 4.18, 3.027 − i0.001, 4.662 − i0.016, 0.64, and 6.36 × 10−6 Å−2,37 respectively. We used monolayer with roughness described by error function to model the brush layer. The thickness, SLD, and roughness were determined by fitting. The thicknesses of chromium and gold layers and the roughness of the gold surface were predetermined by fitting the neutron reflectivity data of bare gold surface before injecting the PEGSH solution in pure D2O measured at three incident angles (0.30°, 0.75°, and 1.80°): wavelength band from 0.2 to 0.88 nm was used. We measured another gold substrate, in which wavelength band from 0.2 to 0.88 nm was used with the incident angle of 0.30°, 0.75°, and 1.80°, which was immersed into 200 μM PEG-SH D2O solution for 40 min and replaced with D2O. The analysis was done in the same way as the first sample. Quartz Crystal Microbalance (QCM). Experiments were carried out using a N2PK Vector Network Analyzer38 operated by QTZ software.39,40 We used AT-cut quartz crystals (Inficon, a diameter of 25.4 mm) with a gold electrode, which have a fundamental frequency of 5 MHz. Quartz substrate was sonicated in ethanol for 15 min and then UV-ozone treated for 1 h before measurement. Temperature was controlled at 30.0 ± 0.1 °C since the resonance frequency of quartz is sensitive to even small temperature changes. In the experiment, quartz substrate was immersed in ultrapure water, and after the resonance frequency became stable, water was replaced with 250 nM PEG-SH aqueous solution. The resonance frequency shift Δf and dissipation
Figure 1. Graft density of PEG brush after the gold substrate was immersed into 250 nM PEG-SH aqueous solution measured by SPR. 0 s is the time when water was replaced with PEG-SH solution. The grafting density grew linearly in the first 5000−6000 s.
increases after the gold substrate was immersed into 250 nM PEG-SH aqueous solution and saturates at a graft density of approximately 0.5 chains/nm2. The grafting rate begins to slow down from linear evolution at 5000−6000 s after gold was immersed into the solution. In this time range, the graft density was 0.20 to 0.23 chains/nm2, so it is suggested that around this graft density the grafting of newly arriving chains is hindered by the already grafted brush chains. SPR experiment quantitatively measures the increasing graft density with high time resolution; however, the structural information such as the thickness of the brush could not be accurately obtained from SPR alone since the brush is very thin (thickness
