In Situ Synthesis of Palladium Nanoparticles in Polymer Brushes Followed by QCM‐D Coupled with Spectroscopic Ellipsometry

The combinatorial method of quartz crystal microbalance coupled with spectroscopic ellipsometry can be used to follow the in situ synthesis of catalytically active palladium nanoparticles in polymer brushes in situ. By combining these two orthogonal techniques, it is possible to thoroughly characterize the thickness, composition, optical, and mechanical properties of nanoscopic thin films. The immobilization of nanoparticles on macroscopic surfaces is of great importance for industrial use, since this facilitates manipulation, stability, and recycling, especially when used as catalysts. Above all, the immobilization of the nanoparticles must not alter or block their surface, so their functionality may be retained. To this end, so-called polymer brushes are regarded as very promising candidate scaffolding materials.1, 2 Polymer brushes consist of polymer chains grafted in close proximity to a surface, forcing the chains to stretch in a so-called brush conformation.3 These systems are capable of responding to external stimuli such as temperature, solvent polarity, and pH, generally by reversible swelling–deswelling behavior. Furthermore, they possess a large number of functional groups that can be exploited to immobilize nanoparticles, while the open structure of the polymer brushes ensures access to the nanoparticles. Not only can these polymer brushes be used to immobilize preformed nanoparticles by interacting with stabilizers on the particle surface4, 5 but the polymer layer is also able to act as a nanoreactor where nanoparticles can be formed in situ.6-8 Recently, we investigated the in situ synthesis of catalytically active palladium and platinum nanoparticles in poly(2-vinylpyridine) (P2VP) brushes, as an efficient way to create catalytically active nanocoatings.9 To control and optimize the properties of these nanocatalysts, it is important to understand the processes involved in the interaction of the precursor with the polymer and the formation of particles inside the layer. By combining the optical method spectroscopic ellipsometry (SE) with the acoustical technique quartz crystal microbalance with dissipation (QCM-D), it is possible to thoroughly characterize the thickness, composition, optical and mechanical properties of polymer brush coatings in situ. Particularly, the mechanisms of adsorption can be tracked and the stimuli–responsive swelling behavior of the polymer brushes can be observed. While SE allows the characterization of thickness and composition of the layer via appropriate optical model approaches, QCM-D gives information about thickness, solvent molecules coupled to the polymer brush and changes in viscoelasticity.10, 11 Applying this hybrid method, we investigated the adsorption of metal ions to P2VP, that are then reduced to Pd nanoparticles in a second step (Figure 1). It had been shown that P2VP interacts with Pd2+ and Pd nanoparticles by formation of a coordination complex via its pyridine groups.9 Figure 2 displays exemplary raw data taken during the adsorption of Pd ions to P2VP brushes. After an initial rinse with a constant flow of ethanol, a PdCl2 solution is introduced (indicated by the shaded area) causing a change in both QCM-D and SE measurements. Comparative measurements with bare quartz sensors demonstrate that this change is only caused by the interaction of the polymer with the solute and not due to changes in the optical and viscoelastic properties of the solution. Experimental SE data acquired in the visible spectral region (400–1000 nm) were modeled with a stratified layer optical model comprising three layers, which account for the substrate, the anchoring layer, and the polymer brush, respectively. The transparent polymer brush layer was described with a Cauchy dispersion line shape. Figure 3 shows the best-match model thickness and refractive index of the polymer brush layer. The P2VP chains do not swell in the solvent ethanol. Upon introduction of PdCl2, the refractive index increases rapidly due to the incorporation of Pd2+ ions with a high refractive index into the polymer layer. After an initial increase, the thickness parameter decreases, this observation may be associated with cross-linking of the polymer chains by the interaction of two pyridine groups with one Pd2+ ion. Upon further incorporation of Pd2+, the thickness parameter slightly recovers, possibly due to electrostatic repulsion of the positively charged moieties immobilized in the otherwise neutral polymer layer. Final rinsing with ethanol causes very slight changes in refractive index and thickness, implying that the Pd2+ ions are tightly bound to the polymer chains and cannot be rinsed off. To quantify the amount of Pd in the polymer layer in the final adsorption stage, an isotropic Bruggemann-effective medium approach (EMA) was used.12 A volume percentage of (5.7 ± 0.1%) Pd was determined, which corresponds to an areal mass of (5.3 ± 0.4 mg m−2) (assuming a bulk density δPd,bulk = 11.99 g cm−3). To model QCM-D data, two approaches had to be used. In the first part of the adsorption, the dissipation decreases, implying decreasing layer viscoelasticity due to a loss of mechanically coupled solvent. Because the dissipation parameter is sufficiently large and the polymer brush layer may be considered laterally homogenous, a Voigt-Voinova model was used.13, 14 In the second part, a simple Sauerbrey model was sufficient to describe the change in frequency.14, 15 Figure 4 displays the change in areal mass, viscosity, and shear modulus fitted from QCM-D data. Similar to the thickness change seen in the SE data, QCM-D shows an initial increase of coupled mass followed by a rapid decrease, which can be associated with the initial adsorption of ions, followed by cross-linking, and deswelling of the polymer layer accompanied by the loss of counter ions and solvent molecules previously co-adsorbed to the polymer layer. Deswelling also causes a drastic decrease in viscoelasticity. The second part of the adsorption takes place without further change in viscoelasticity, and the further incorporation of Pd2+ ions can be monitored by the change in adsorbed mass only. In agreement with SE measurements, the final rinsing step causes only a slight decrease in the adsorbed mass, again implying the tight adsorption of Pd2+ ions to the pyridine group of the polymer. A final mass change of ≈2 mg m−2 is calculated. Thus, the mass change resulting from QCM-D and SE measurement analysis is comparable, although both values have to be regarded with care. The mass change of QCM-D was gained by the Sauerbrey equation neglecting the viscosity and the density of the ambient. For SE on the other hand, the assumption of the optical properties and density of bulk Pd for the adsorbed ions leads to an overestimation of the areal mass. Nevertheless, that both techniques show comparable results, provides some assurance that both assumptions are reasonable. Due to the formation of hydrogen bubbles, the reduction to Pd nanoparticles with NaBH4 could not be monitored in situ. Figure 5a compares Ψ(λ) and Δ(λ) of SE measurements taken at the three steps of the nanoparticle formation. No further variation of the ellipsometric angles is detected after the reduction of Pd2+ to nanoparticles. Analysis of the measured data results in almost no change in Pd content and layer thickness, which suggests that all Pd2+ ions have been reduced to Pd nanoparticles (Figure 5b). Figure 6 displays atomic force microscopy (AFM) height images of P2VP brushes on a reference silicon wafer before and after the adsorption and reduction of Pd2+. It can be seen that nanoscopic particles have been incorporated in the polymer layer in a homogeneous manner. In conclusion, the combinatorial approach of QCM-D and SE proves to be a valuable tool to investigate the adsorption of metal precursors to a thin polymer layer for the formation of nanocatalytic coatings. With relatively simple data analysis procedures, it is possible to gain quantitative information about the amount of metal inside the polymer layer and changes in brush conformation and viscoelasticity during the adsorption process, leading to new insights in the process of metal ions adsorbing to a neutral polymer layer. Since QCM-D and SE data were taken simultaneously, it is possible to directly compare the results of these orthogonal methods for cross-validation. Both methods show an initial Pd2+ uptake followed by polymer chain cross-linking, the latter process causing a decrease of thickness. In the last step, Pd2+ is further incorporated into the now more rigid polymer layer. Measurements during rinsing steps and after the reduction of Pd2+ to Pd nanoparticles demonstrate that the ions are tightly bound to the polymer and all ions are reduced to nanoparticles. The problems that arose due to bubble formation upon reduction in NaBH4 solution could be avoided in the future, by use of an open fluid cell, developed by Richter and co-workers.16 This combinatorial setup offers the possibility to comprehensively investigate the properties of nanoscopic thin films in situ and is therefore of interest for various investigations involving nanotechnologies. Both QCM-D and SE can be operated under various ambient conditions with a high temporal resolution, thus allowing the close tracking of processes occurring at the surface, which is less available by other surface-sensitive methods. Since both techniques can be operated on a variety of different substrates and over a large thickness range, the combinatorial setup is a versatile tool to analyze films for both organic or biological applications, such as self-assembled monolayers, swelling of polymer films or adsorption of biomolecules, and inorganic applications, such as nanoparticle adsorption or metal deposition. Materials: Poly(glycidyl methacrylate) (PGMA; = 17 500 g mol−1, PDI = 1.7) and carboxy-terminated poly(2-vinylpyridine) ( = 40 600 g mol−1, PDI = 1.08) were purchased from Polymer Source, Inc., Canada. Palladium chloride (PdCl2) was purchased from Sigma–Aldrich, Germany. Sodium borohydrate (NaBH4) and chloroform (CHCl3) were purchased from Sigma–Aldrich, USA. 200-Proof ethanol was purchased from Pharmco, USA. Silicon-coated quartz crystals (QSX 303, Q-Sense, Frölunda, Sweden) were used as substrates. Purified water was used throughout the experiments. Instrumentation: Within the QCM-D/SE setup, a M-2000 ellipsometer (J. A. Woollam Co. Inc., Lincoln, NE, USA) with a spectral range from 400 to 1000 nm and an ellipsometry-compatible QCM-D module (QELM 401, Q-Sense, Frölunda, Sweden) with a fixed angle of incidence of 65° was used. AFM images were taken with a Dimension Icon microscope (Bruker Corporation, Karlsruhe, Germany) in peak force mode. A first-order plane-fit was applied to the raw data images with NanoScope Analysis Software (Bruker Corporation, Karlsruhe, Germany). A SPI Plasma Prep II plasma cleaner (Structure Probe, Inc., West Chester, PA, USA) was used for cleaning and activation of substrates. Preparation of Polymer Brushes: P2VP brushes were prepared by the already optimized “grafting to” approach.17 Silica-coated quartz sensors were first cleaned by rinsing with ethanol abs. and activated in oxygen plasma (1 min at 100 W). A layer of PGMA was spin coated onto the wafers from a solution in chloroform (0.02 wt%) and then annealed (20 min at 100 °C). Onto that anchoring layer, the P2VP chains were grafted by spin coating from a solution in chloroform (1 wt%) and annealing (3 h at 150 °C). Ungrafted polymer was extracted in 200-proof ethanol. Each step of the process was controlled by an ellipsometry measurement of the samples in the dry state. Synthesis of Pd Nanoparticles: Experiments were done under flowing conditions (1 mL s−1). After an initial ethanol rinse, sensors coated with P2VP brushes were rinsed with a solution of PdCl2 in ethanol (0.14 × 10−3 m). To remove loosely adsorbed metal ions, samples were subsequently rinsed with ethanol. For the reduction to nanoparticles, samples were rinsed with a NaBH4 solution (0.2 m) and finally rinsed again with water and ethanol. Ellipsometry Modeling: Measurements of the blank silica-coated quartz sensors were parameterized by a B-Spline model and taken as a virtual substrate. For dry measurements, the modification with polymer layers was modeled with a fixed refractive index (nPGMA = 1.525, nP2VP = 1.595) in an optical box model. For in situ measurements, a two-parameter Cauchy dispersion model was used to determine both refractive index and thickness. The volume fraction of Pd was determined by a three-component Bruggemann EMA model that used the P2VP index of refraction experimentally determined earlier and literature values of the Pd dielectric function.18 The optical properties of the ambient ethanol were measured by using a VUV-VASE and the minimum deviation approach.19 QCM-D Modeling: Shifts in frequency and dissipation of the odd overtones (j = 3,5,7,9,11) with reference to the measurement with the smallest dissipation value were modeled by either a Voigt-Voinova model for one homogeneous viscoelastic layer with a fixed density of 1 g cm−3 or the Sauerbrey relation using the software QTools (Q-Sense, Frölunda, Sweden). Financial support was granted by the German Science Foundation (DFG) within the DFG-NSF cooperation project (DFG Proj. Nr. STA 324/49–1 and EI 317/6–1) in the frame of the “Materials World Network,” in the frame of the priority program SPP 1369 “Polymer-Solid Contacts: Interfaces and Interphases” (DFG Proj. Nr. STA 324/37–1), and the National Science Foundation within the NSF-EPSCoR project (Proj. Nr. 1004094).

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.