Ordering of Nanoparticles by Wrinkle-Assisted Self-Assembly : Controlling Plasmonic Coupling Effects
- Structures of spatial scale between 10Å and 1000Å are known as nanomaterials and have attracted immense interest over the last decades (Nobel Prize in physics in 2010 was awarded for the nanomaterial graphene). Materials within this scale show a large surface-to-volume ratio and amplify surface-related properties. Governing and manipulating material on this almost atomic level is one of the most active fields in modern natural science. Nanoscale technology, such as some of the processes involved in steel production and painting, has been empirically utilized in human society for centuries, however, a scientific investigation of phenomena on this spatial scale only began in 1857 when Michael Faraday reported on the synthesis and colors of gold colloids. In 1959 interest in the nanoscale was stimulated by an American physicist, Richard Feynman, in his famous “There’s plenty of room at the bottom” address, and the term nanotechnology first appeared in 1974 from the Japanese Norio Tanigucho. Since these pioneering works, thousands of publications have been focused on the synthesis, modification, properties and assembly of nanoparticles. Great progress has been attained in the preparation of nanoparticles of any desired size, shape and composition. Metal nanoparticles are particularly attractive due to their spectacular size and shape dependent optical and electronic properties. Color variations of nanoparticle suspension for example arise from changes in the composition, size and shape of nanoparticles, as well as from the proximity of other metal nanoparticles. The average distances of nanoparticles in thin films influence the spectral features because of inter-nanoparticle coupling. These effects are often the result of changes in the so-called surface Plasmon resonance, the frequency at which conduction electrons oscillate in response to the alternating electric field. Provided nanoparticles form ordered arrays, they can additionally have unique and fascinating optical properties because of photonic band gap effects with potential applications such as detectors, circuits, light sources, polymeric opals or meta-materials. The present work deals with the controlled placement of nanoparticles by physical con-straints. Exact placement of nanoparticles allows for the control of the inter-nanoparticle distance and thus determines the coupling effects (here: Plasmon coupling) which arise upon interaction with electromagnetic radiation. Different coupling leads to different distance-dependent signals and such substrates can serve as sensors if, for example, Raman spectroscopy is carried out for detection of the signal. Currently, most templates are created using lithographic techniques. Particularly if structures on the sub-micron scale are desired, electron beam lithography has to be used which involves environmentally harmful etching processes. Within this work we show how controlled wrinkling of a thin rigid film on a soft, elastomeric substrate, can be used as an alternative to fabricate nano-templates without using any lithography. As a substrate, a silicon elastomer poly (dimethylsiloxane) (PDMS) was used. Upon stretching such substrates uniaxially, an enlarged surface was exposed to oxygen plasma and converted to silica by oxidation. After releasing the strain, periodic wrinkles appeared perpendicular to the applied strain. Under defined conditions, such wrinkles have a regular sinusoidal topology featuring a single dominant wavelength and amplitude. The formation process could easily be tuned by tuning the plasma exposure to generate periodically structured templates between few hundreds of nanometers and several microns. In this work, wrinkled templates were tailored such that suitably sized nanoparticles could be arbitrarily assembled into a hierarchical structure by drying colloids out of suspension in a channel-like confinement offered by wrinkles in contact with a flat substrate. Using the same template geometry (same wavelength and amplitude of wrinkles) but different particle concentration of spherical polystyrene beads (r = 55nm) we found parallel particle-structures ranging from single parallel lines at low particle concentration to dense prismatic ridges at high particle concentration. The wavelength of the wrinkled template defined the spacing between the particle lines. Moreover, we performed Monte Carlo (MC) computer simulations in collaboration with the theoretical physics department (Prof. Dr. Matthias Schmidt and Dr. Andrea Fortini) at the Uni-versity of Bayreuth to assess the dominant driving forces during the assembly process. Be using MC, colloidal particle assemblies can be characterized in terms of their equilibrium configuration that minimizes the free energy. Simulations were performed on particles in a box delimited by a flat hard wall and a sinusoidal hard wall according to our experimental system. These simulations precisely predicted the exact assembled geometry in thermal equilibrium. Comparing results of simulation and experiment we found perfect agreement between the equilibrium structures. We discovered the confinement itself to be mainly responsible for the assembled morphology of nanoparticle, which makes the process independent of the detailed chemistry of particles. In addition we obtained very similar structures with the same assembly strategy but using gold nanoparticles (r = 33 nm) instead of polymeric particles. We fabricated lines of gold nanoparticles assembled in a single file and lines two particles wide using similar particle concentration but different sizes of the confinement template. The different morphologies of the lines give rise to different optical signals as collective oscillation of conduction electrons result in different interaction with electromagnetic radiation. Surface Plasmon resonance due to Plasmon coupling between adjacent particles arises. Different morphology-dependent signals of nanoparticles in contact within the lines were detected by surface enhanced Raman spectroscopy (SERS). The electromagnetic field was measured to be randomly distributed along the particle lines with strong enhancements at so-called hot spots located at gaps between neighboring nano-particles. To confirm the measured signal we compared theoretical simulations using the finite-difference time-domain (FDTD) method and experimentally measured dark-field spectroscopy signal along differently shaped lines of particles within a collaboration with Weihai Ni and Dr. Ramón Alvarez-Puebla at the University of Vigo in Spain. Good agreement between theory and experiment indicated that indeed plasmonic coupling of the individual nanoparticles is responsible for the observed SERS effects: Using wrinkle-assisted self-assembly it is possible to control the organization of the colloidal particles on the substrate, with a consequent control over the formation of hot spots and the resulting SERS intensity. Such ordered multiplicities of hot spots give rise to quantitative SERS signals with high sensitivity which has applications as diverse as biological detectors, optical filters and sensors. In addition, this work deals with chemical modification of the wrinkled structure to render it accessible to different solvents as PDMS tends to swell in organic solvents and suffers from poor mechanical stability. Additionally, wrinkles fabricated through a buckling instability of a stiff supported layer under compression are not tension free on the microscopic level and suffer from relaxation on a longer time scale. We introduce in this work two different methods to replicate wrinkles by molding. In micro thermoforming, the wrinkled surface was used as a mold (or caliber) to structure different kinds of polymers (polystyrene and poly (methylmethacrylate)) by pressing the originally wrinkled structure onto a ductile material which preserves the nanostructure after curing. The second methodology was carried out in collaboration with PD. Dr. Kerstin Koch and Michael Bennemann at the Nees institute in Bonn and employed a two-step molding process, where wrinkles were molded against wax and in a second step, the structured wax was cast against epoxy resin. Both methods revealed perfect copies of the wrinkled original with high fidelity even at dimensions as small as a few hundred nanometers and hold no residual stresses because there is only one component. Wrinkles made of tough polymers are now accessible to various solvents which make them potential substrates for microfluidics. In the last part of this work, wrinkles are used as stamps in so-called micro contact printing (µCP). In this technique, a structured elastomeric stamp is used to transfer a surface-active molecule out of solution to a flat substrate by mechanical contact. Patterns of different charge density can be created which have applications in the field of biosensors, diagnostic immunoassays and cell culturing. Traditionally, stamps for µCP are prepared by a two step process where a lithographically fabricated structured silicon master serves as mold. An elastomeric polymer is cast against the caliber and preserves the structure after curing and detaching. As already mentioned lithography is expensive and involves environmentally harmful etching processes. Within this work we introduce the one step wrinkling process to fabricate structured stamps. Even though the diversity of stamp geometries created by wrinkling is limited, the simplicity compared to lithographic techniques is evident. The process of wrinkle formation includes plasma oxidation, which renders the topmost surface hydrophilic. Therefore, charged macromolecules out of aqueous solution were adsorbed onto the surface. The coated relief structure was used as a stamp to transfer the molecules selectively from the elevated parts of the wrinkles to another flat, oppositely charged surface by means of µCP. The topography of the resulting pattern was characterized by Atomic Force Microscopy (AFM) imaging as alternating charged pattern of printed (elevated) and non-printed areas. By varying the geometry of the wrinkled stamp (amplitude and wavelength) we studied the limits in which successful µCP with wrinkles can be carried out. We found the limits for wavelength of the wrinkles below 355nm and amplitudes below 40nm at which the printed structure disappeared because material was transferred from the wrinkles’ hills as well as from the bottom parts. The height of the transferred structure increase with increasing wavelength and amplitude of the wrinkles but tended to a limit of 6-7nm, even though the topology of the stamp increases. The smallest structure found in lateral dimensions was as small as 50nm, appearing as areas where no material was transferred.