Plasmonic nanoantennas for multipurpose particle manipulation and enhanced optical magnetism

Optical manipulation – using light to control matter – is based on the transfer of momentum from confined electromagnetic fields to micro- and nanoscale objects. Optical tweezers, based on the use of high-numerical aperture objective lenses, take advantage of this phenomenon to form non-contact probes that are capable of applying piconewton forces and detecting motion with angstrom-level precision. While the technique originated essentially as a scientific curiosity in the 1980s, it has since revolutionized the field of single-molecule biophysics, provided key insights into motor-protein function and DNA structure, and enabled the formation of the first Bose-Einstein condensates, and is the subject of significant current interest for lab-on-chip, colloidal physics, and biochemical applications. Despite the clear significance of this technology, optical tweezers are constrained by the diffraction limit, which places an upper bound on achievable optical forces in a given system. This limitation is problematic as demand for investigating increasingly smaller, nanoscale systems is on the rise. An alternative approach based on the subdiffraction field confinement and enhancement properties of metallic nanostructures is a promising avenue to circumvent the problem of diffraction-limited forces. The near-field intensity gradients produced in the nanometer-sized gaps of plasmonic nanoantennas are orders of magnitude larger than those of conventional optical tweezers. Accordingly, plasmon-enhanced gradient forces can both significantly relax constraints for microparticle manipulation and offer a route for improved nanoparticle trapping. In addition, resonant optical absorption in plasmonic devices leads to considerable heat generation, which in turn can induce convective flow in the fluid environment containing the particles. While previous studies have regarded this process as deleterious to the performance of plasmonic optical tweezers, careful selection of optical and geometric parameters can make plasmon-induced convection favorable for particle manipulation. This dissertation explores the near-field enhancement and confinement properties of arrays of Au bowtie nanoantennas (BNAs) for plasmonic optical trapping. Using BNAs as a model system, the delicate interplay between optical and thermally induced forces in plasmonic nanotweezers is investigated over a broad parameter spacing including bowtie array spacing, adhesion layer materials, nanostructure orientation, composition of the fluid trapping media, optical polarization, input optical power, and trapped-particle diameter. Using theoretical modeling, it is shown that plasmon-induced convection drives experimentally observed phase-like behavior in plasmonic nanotweezers, and further, that this process can be used to engineer trapping tasks including dexterous single-particle trapping, trapping and manipulation of large self-assembled particle clusters using a single input beam, and particle sorting. The crucial role of an optically-absorptive substrate material for developing the requisite micron-per-second fluid flows for these phenomena is confirmed both theoretically and experimentally. In addition, this dissertation details the first experimental demonstration of plasmonic nanotweezers using an ultrafast, femtosecond (fs) pulsed input source. The fs pulses are shown to increase trapping performance in both the Rayleigh and Mie size regimes, where particle diameters are much smaller and greater than the incident illumination wavelength, respectively. This augmentation of forces enables plasmonic trapping of 80 nm to 1.2 m diameter, metallic and dielectric particles with as little as 50 W of input optical power. Moreover, the nonlinear optical response of trapped species can be probed during the trapping event, which opens doors for increased particle diagnostics in plasmonic optical trapping. An interesting particle fusing behavior is described whereby above a 60–75 W power threshold, both metallic and dielectric particles spontaneously fuse to the BNA surface, likely by means of fs-augmented near-field gradient forces. Using this particle-fusing behavior as inspiration, a novel class of “capped” nanoantennas is designed, and their plasmonic response is theoretically investigated. The specific example of capped-bowtie nanoantennas (c-BNAs) is chosen, and it is shown that the c-BNAs have the unique ability to simultaneously enhance both magnetic and electric fields by more than three and four orders of magnitude, respectively. This ability improves on currently available designs that enhance magnetic fields at the expense of a mitigated electric response. The spectral response of the c-BNAs is dominated by two distinct resonant peaks: one in the visible (VIS) and one in the near-infrared (NIR), and the spectral behavior of the c-BNAs is examined as a function of cap thickness, bowtie gap spacing, and c-BNA array spacing. Finally, a new pillar-bowtie nanoantenna (p-BNA) design, comprising Au BNA arrays suspended on 500 nm tall SiO2 pillars, is introduced as a candidate system to show, for the first time, that the mechanical degree of freedom (DOF) can be used to create in situ reconfigurable plasmonic nanoantennas. Reconfigurability is achieved using electron-beam manipulation in a scanning electron microscope (SEM), whereby the electron beam induces strong electromagnetic gradient forces in the p-BNA gap that causes the two arms to deform toward the common gap center. In characterizing this behavior as a function of SEM accelerating voltage and magnification, design curves are produced that enable controlled, repeatable fabrication of nanoantennas with gap sizes as small as 5 nm by actuation of the mechanical DOF of the pillars. As a proof of this novel design principle, the optical response of two, 10 x 10 modified p-BNA regions comprising 5- and 15-nm gap antennas is characterized using spatially localized reflection spectroscopy based on a supercontinuum optical source