A chemical reaction is fundamentally initiated by the restructuring of a chemical bond. Chemical reactions occur so quickly that their exact trajectory is unknown. To unlock the secret, first one would seek to know the inner working of a single molecule, and therein, a single chemical bond. However, the task is no small feat. Single molecule studies require exquisite spatial resolution afforded by relatively new technologies, and ultrafast laser techniques. The overarching theme of my dissertation will be the path towards achieving the space-time limit in chemistry: namely, the ability to record the structural changes of individual molecules during a reaction, one event at a time.

A scanning tunneling microscope (STM) is used to image the molecules and manipulate their electronic environments. STM has the capacity to create topographical images of molecules with \r{A}ngstr\"{o}m ($10^{-10}$ m -- the size of an atom) resolution, and can also probe the molecule electronically by use of a tunneling current ($I_t$). STM images reflect the changes in the potential energy surface (PES), and help us understand how molecules interact with surfaces and each other, thereby accessing the fundamental problem of catalysis and chemical reactions. In addition to seeing the molecule, we use Raman spectroscopy to track its molecular changes with chemical specificity. I combine these experimental tools to investigate tip-enhanced Raman spectra (TERS) of single molecules within the confines of a STM.

These methods were used to report the conformational change of a single azobenzene-thiol derivative molecule. Although we were able to definitely isolate a single molecule signature, imaging the single molecule in real space and time proved elusive. Additionally, I report on a conductance switch based on the observable change of the topographic STM images of a radical anion mediated by the spin flip of a single electron on a single molecule. This is effectively the smallest achievable architecture of molecular electronics, negating the need for heat dissipation in small systems. A related work found how physisorption potentials of molecules to metals could be experimentally visually verified and modeled by STM, thus allowing us to use the STM tip as a driver for molecular motion on surfaces.

Throughtout this work, we noted that a dominant feature of single molecule chemistry are intensity and spectral fluctuations that are difficult to characterize, as the molecule contorts wildly when it experiences distinct and powerful electromagnetic fields and field gradients. This much is evident in the last experiment, and chapter, of this thesis. Raman spectra associated with cobalt (II) tetraphenyl porphyrin (CoTPP) axially coordinated with bipyridyl ethylene (BPE) were captured with Raman mapping with nanometer resolution. However, the stochastic apperance of Raman lines and low resolution images made it difficult to ascertain which molecule we captured. The preliminary results as well as follow-up conrol experiments are discussed.

While each experiment constitutes in and of itself an important, individual contribution, their sum establishes the principles of seeing single-molecule chemistry.

Time-resolved, surface-enhanced, coherent anti-Stokes Raman spectroscopy (tr-SECARS) is

ideally suited for preparing and interrogating vibrational coherences on single molecules.

We have succeeded in the rst demonstration of this concept through measurements carried

out on molecules attached to gold nanosphere pairs which act as plasmonic nano-dumbell

antennae. The tr-SECARS traces provide unique signatures of coherent evolution in discrete

ensembles. The signals are characterized by phase and amplitude noise, which can be cast in

terms of amplitude probability distribution functions (PDF), which allow rigorous distinction

between single, few, and many molecule coherences. We give a brief background on tr-CARS,

the experimental system for carrying out tr-SECARS and the analysis of the results in terms

of PDFs. The analysis makes it clear that we have, for the rst time, observed the coherent

vibrational motion of a single molecule.

Optical activity has long been an area of interest in Raman scattering spectroscopy as it is a powerful tool for elucidating molecular chirality. Here, I provide details from an investigation into plasmonic (rather than molecular) optical activity in surface-enhanced Raman spectroscopy. The sample used is the prototypical dumbbell nano-antenna (nantenna) consisting of two gold nanospheres functionalized with bipyridyl ethylene molecular reporters. Previous analysis of the linear dichroism reveal that the structures scatter as dipolar antennas, with the molecular vibrations following the polarization patterns of the dumbbell on which they reside. Current investigations into circular optical activity of the nantenna reveal two notable observations. First, the Raman optical activity is at least two orders of magnitude larger than values typically observed for molecular species. Second, the observed handedness of the dumbbell is entirely dependent on its orientation in the plane perpendicular to excitation.

We attribute this counter-intuitive effect to the multipolar response of the plasmonic nantenna, which arises from its large size with respect to the excitation wavelength (~λ/2). Here, the long-wave approximation breaks down and retardation effects can no longer be ignored, giving rise to the inclusion of the electric quadrupole and magnetic dipole modes in addition to the zeroth-order electric dipole. The observed ODH can be understood in terms of the electric and magnetic dipole terms, which appear to undergo a simultaneous time reversal operation upon in-plane rotation. These results are explained in the framework of Jones calculus and have important implications in the treatment of bi-isotropic media.

Interactions between photons and plasmons have resulted in a hotbed of research activity in the physical sciences, notably through techniques such as white-light and Raman scattering, as evidenced by >10,000 articles published in 2020. In this work, plasmon-mediated linear white-light scattering and nonlinear spontaneous and coherent Raman scattering techniques are explored. In the first part of the thesis, I present studies of white-light scattering from nanostructures assembled using an in-situ scanning electron microscope (SEM)-compatible nano-manipulator robot. I discuss the challenges and derive the limitations of manipulating nano-sized objects using the nanorobot, and address geometry considerations related to white-light scattering from the nanostructures. In the second part of the thesis, I present studies of Raman scattering of bipyridyl ethylene (BPE) trapped at plasmonic nanojunctions, for historical reasons, known as surface-enhanced Raman scattering (SERS). We observe anomalously large anti-Stokes-to-Stokes scattering ratios accompanied by enhancement factors on the order of 1011 - 1012. We show that the effect does not arise from sequential Raman excitation; instead, it results from single-beam coherent anti-Stokes Raman scattering (CARS) where the Stokes plasmon acts as the intracavity stimulating field. I present conventional two-beam surface-enhanced CARS (SE-CARS) measurements on the same system using picosecond lasers, in both frequency and time-domain, and show that the single beam limit can be reached by attenuating the incident extracavity Stokes field.

Time-resolved, surface-enhanced, coherent anti-Stokes Raman spectroscopy (tr-SECARS) is

ideally suited for preparing and probing vibrational coherences in molecules. By enhancing

the local response of a single molecule with a dipolar nano-antenna, vibrational dynamics

have been measured at the single molecule limit. In contrast with tr-CARS measurements

in ensembles, the vibrational coherence of a single molecule is not subject to pure dephasing.

It exhibits characteristic phase and amplitude noise, which allows the statistical distinction

between single, few, and many molecule sources to be determined. To build on the cur-

rent work, by using three unique pulses to spectrally lter the response of the molecule,

the characteristic noise can be isolated and measured background-free. If the probing of a

superposition state is carried out over a real resonance, then it is possible to tomographically

reconstruct the complete description of quantum dynamics in phase space representation via

the Wigner Distribution Function(WDF). The WDF can be reconstructed from either the

wavepacket via Wigner Transform, or an experimentally measured density, via an Inverse

Radon Transform. The calculations presented here highlight the necessary conditions in

order to reconstruct the WDF with delity from a proposed experiment and compare the

density derived WDF with that of the wavepacket. The principle is rstly demonstrated us-

ing a Kerr gated detection of emission from an evolving state on a bound harmonic potential

energy surface. The model is then explained in the case of a proposed spectrally resolved

transient grating experiment (SRTG). The WDFs generated from the limiting conditions

show that the reproduction delity of the experimentally derived WDF are dependent on

the probe, utilized to measure the evolving superposition, and the curvature, or the vibra-

tional frequency of the potential energy surfaces, and the dephasing time of the vibrational

superposition states. Given two potentials, I show that it is possible to optimize probe

pulse parameters to improve the delity of the state reconstruction. Due to the variational

principle, the negative volume of the WDF, or the Wigner hole, can only be reduced via

measurements - the pulse parameters can be optimized iteratively even when the molecular

potentials are not known.

The investigation we present here of SPP driven field emission, after propagation to a probe tip axis, leads to two firm conclusions. First, we have demonstrated nonlocal electron emission into vacuum driven by an SPP coupled to a surface after geometric focusing. Emitted electrons were measured as a current either using a weak (emission rate independent) collecting bias or with a strong DC voltage whose apex focused local field contributes to the work function suppression of the SPP driven field emission. Second, we demonstrate a number of informative effects of this emission behavior through comparative studies of material, applied voltage ranges, and a fully optimized and controlled surface geometry. The most important of these findings were thermal effects and local damage for this unique case of emission from sites far from incident radiation. While these observations demonstrate efficient coupling to SPPs, they underscore the limitations of local field strength attainable with propagating modes.

Surface-enhanced Raman spectroscopy (SERS) has been an area of active research since its discovery in 1973. In particular, the dimer nano-antenna SERS geometry, consisting of two gold nanospheres functionalized with molecular reporters, has emerged as a powerful and feasible tool for single molecule studies. Despite its popularity, there is still little consensus regrading the role of the nantenna in dictating the observed SERS response. Until issues such as SERS backgrounds, anomalous molecular anti-Stokes intensities, and polarized responses are fully understood, the full power of single molecule Raman can never be realized.

Here, the antenna behavior is clarified via two types of studies: one excitation-dependent study, in which the SERS signal is recorded as a function of increasing laser power, and one polarization-resolved study, in which the SERS response is monitored as a function of changing excitation polarization. Two important discoveries emerge clearly here: first, that the background commonly observed in SERS spectra can be unilaterally assigned to the inelastic light scattering (i.e., Raman) of the plasmon. Second, that the dimer alone dictates the polarization of the scattered response, including a clear demonstration of both linear and circular chirality. With these in hand, novel applications, such as a quantum description of the plasmon and orientation-dependent chiral response are explored.

Surface-enhanced vibrational spectroscopy has emerged as a powerful tool to probe the (photo)chemistry and (photo)physics of individual molecules, which may be regarded as the primary focus of the field of nanophotonics. Single molecule (SM) sensitivity relies on the excitation of localized surface plasmons which confine far-field radiation to molecular length scales, ultimately enabling submolecular imaging with chemical selectivity. Over the length scales that enable submolecular resolution, localized optical fields cannot be separated from the charge density oscillations that sustain them. As such, the interplay between plasmons, photons and electrons in plasmonic nano-junctions is of critical importance for the development of imaging modalities capable of circumventing the diffraction limit and, more generally, for gaining broader insight into the fundamental mechanisms of surface-enhanced Raman and tip-enhanced Raman scattering (SERS and TERS).

In what follows, the SM surface-enhanced parameter space is explored through linear and ultrafast optical measurements utilizing nanometer to ångstrom scale plasmonic junctions and non-resonant molecular reporters. SERS measurements carried out on the prototypical gold dimer nano-antenna highlight the interplay between the molecular Raman scattering channel(s) and coherent luminescence of the plasmon itself. Ultrafast studies utilizing 100 femtosecond laser pulses establish the operating principles of the ultrafast analog of surface-enhanced spectroscopy, which are then utilized to demonstrated time and frequency resolved coherent Raman scattering (CRS) in the single molecule limit.

The fundamental mechanism of SERS is explored by replacing the continuous-wave excitation source with a picosecond pulse train, which probes the non-equilibrium population dynamics of SM phonon states. The latter are inferred through the observed inverted Stokes/anti-Stokes ratios and verified through stimulated and coherent Raman scattering (SRS and CARS) measurements carried out on the same system. Direct plasmon-molecule energy transfer is dominant over the optically induced and thermal contributions to the vibrational population.

Finally, SM TERS measurements utilizing a metalloporphyrin functionalized silver tip are carried out in ultrahigh vacuum and ultralow temperatures (5 K). These measurements directly image the atomic lattice of a copper nitride monolayer with 1.5 Å precision. A series of measurements recorded on this system highlight the essential role of photoelectrons in the TERS process.

Field emission takes place upon focusing propagating surface plasmon polaritons (SPP) at the apex of a sharp metal tip. The effect is demonstrated with remotely launched SPPs on silver and gold probe tips. We couple femtosecond laser pulses through a grating inscribed on the taper of a smooth, silver wire 30um from the apex. Field-emitted current is directly correlated with the radiation of the super-focused SPP at the apex. Both current and radiation at the apex are measured as a function of incident polarization on the grating. In the absence of incident light at the apex, the local field of the "naked" surface plasmon modulates the tunneling barrier that drives the field emission. We give a detailed analysis of the governing dynamics in the presence and absence of an applied extractor field, and aim to distinguish between mechanisms governing measured current. Independent of the distribution of the electrons in the metal half-space the emission acquires coherence by the time-dependent field of the plasmon in the vacuum half-space.

Experimental techniques with the sensitivity to measure properties of individual molecules present an opportunity to investigate and understand fundamental chemical and physical processes. One such technique, scanning tunneling microscopy (STM), has proven itself to be the best tool for probing the properties of single molecules, enabling researchers to measure a molecule’s electronic and vibrational properties with angstrom-level precision. Combining STM with optical vibrational spectroscopy allows the technique to measure molecular properties, bonding dynamics, optical phenomena, and light-matter interactions that are invisible to bare STM. Developing and perfecting the coupling of lasers with the STM tunneling junction can lead to time-resolving the technique, enabling sub-molecular topographic imaging with femtosecond time resolution. In this work, I detail the results of single molecule investigations with STM. In particular, I highlight the progress I have made in improving ultra-high vacuum STM by combining it with tip-enhanced Raman spectroscopy (TERS). The improvements made to TERS-STM focus on a variety experimental considerations including optimizing laser-tip beam alignment through ray-tracing simulations and drastically improving the performance and reliability of TERS tips through the introduction of new tip manufacturing and preparation techniques. The efficacy of these improvements are demonstrated by the many high-resolution STM measurements including (1) the discovery and characterization of a single-election switch involving the spin flip of an unpaired electron in a single molecule, (2) the observation via TERS-STM of one molecule undergoing a single reaction as well as an analysis of intensity and spectral fluctuations common to TERS, and (3) the demonstration and analysis of angstrom-resolved TERS-STM and its significance to our understanding of single-molecule physics and the tip-enhanced Raman effect.