The basic understanding of chemical reactions relies on the microscopic details of bond breakage and formation, yet the lifetimes of intermediate species involved in chemical reactions are often too short for conventional characterization. Femtosecond infrared spectroscopy is thus a valuable experimental tool for directly identifying the sequential steps of reactions in solution at ambient temperatures. The liquid side of the Harris group uses ultrafast UV/visible-pump, IR-probe spectroscopy to access these fast chemical dynamics and gain new insights in a variety of model photochemical reactions important for catalytic and synthetic applications. In addition to femtosecond vibrational spectroscopy, the liquid side uses modern computation methods, including ab initio calculations, extensive DFT calculations, and molecular dynamics simulations, to understand and interpret the chemical dynamics. While these experimental and theoretical methods have been extremely successful, the liquid side is currently adding the capability for transient two-dimensional infrared spectroscopy (T2D-IR). This experiment potentially offers a more detailed level of information on chemical reactions, including evidence for the transformation of vibrations from reactant to product, molecular anharmonicities, solvation mechanisms, and transient molecular structure. This new technique, coupled with our theoretical approach, should provide detailed information on the dynamics of barrier crossing, ultrafast molecular rearrangments, and solvent reorganization, among others.
The output of a 1 W regneratively amplified Ti:Sapphire laser system is split to pump both a UV/visible pump line and a mid-IR OPA. The time delay between the pump and probe is determined with a computer controlled delay stage. After the OPA, the mid-IR line is split to provide a mid-IR narrowband pump pulse and a broadband IR probe pulse. The frequency of the narrowband pump pulse is determined by a feedback controlled Fabry-Perot filter which narrows the frequency of the IR from ~200 cm-1 to ~15 cm-1. The possible pulse sequences with this apparatus are illustrated in the figure above. By scanning the frequency of the mid-IR pump and spectrally resolving the broadband IR probe in the spectrograph, a two dimensional infrared spectrum can be generated. An example one dimensional spectrum is shown below.
The spectrum at one time delay in the figure above can be collected with less than one minute of data collection. The data are presented as difference spectra, which means absorptions depleted by the UV/visible pump pulse result in negative peaks, while new absorption formed after the pump pulse result in positive absorptions. Typically, the UV/visible pulse is used to initiate a chemical reaction, such as CO photodissociation or homolysis of a metal-metal bond in organometallic complexes:
The reactions occuring after photoexcitation are monitored by probing the stretching frequencies of CO groups attached to the metal centers. Small changes in the electron density on the metal center causes significant changes in the vibrational frequency. For example, the spectrum shown in the figure above displays absorptions from five distinct chemical species. The changes in the absorptions as a function of time provides detailed information on the chemical dynamics.
Ab initio, density functional theory calculations, and molecular dynamics (MD) simulations are important computation methods used on the liquid side to gain detailed insight into the chemical dynamics we observe with our femtosecond laser system. These computation methods are typically performed in house, providing a strong background for students in both experimental and theoretical methods. Ab initio calculations are used to identify previously unobserved (because they are short-lived) chemical species and to identify the transitions states and barriers, which are directly related to the chemical dynamics we observe experimentally. MD simulations are used in cases where the solvent dynamics are extremely important to understand the behavior of the system. For example, we recently used transition path sampling MD to understand the dynamics of ligand exchange.
A variety of projects, including the dynamics of beta-H elimination, the reactivity of 17-electron and 19-electron radicals, rearrangements of metal clusters, and dynamics of solvent rearrangment upon CO photodissociation, among others, are currently in progress or have recently been completed. Please view our recent publications for further details on these projects or feel free to contact us about these or new projects.
The study of the electronic properties of interfaces between dissimilar materials constitutes a major research area of fundamental interest and technological significance. In principle, chemical and physical processes occurring at interfaces can be understood on a microscopic level in terms of the associated energy levels and dynamics of the interfacial electrons. A molecular scale understanding of interfacial electronic properties should help facilitate the rational design of photovoltaic cells, molecular electronics, and heterogeneous catalysis, to mention a few areas of current interest. Our goal is to understand the response of interfacial and surface electronic states in condensed matter after an electron has been injected.
We use two-photon photo-emission (2PPE) to examine the fundamental physics and chemistry of electronic structure and electron dynamics on the femtosecond (10-15 sec) time scale, at interfaces between a single crystal metal substrate and molecular films a few monolayers thick. Recent adsorbates under investigation have included simple polar molecules, which are models of electrochemical solvents and their solvation/localization behavior; and large aromatic molecules, which are model molecular electronics. As the sizes of electronic devices and circuit elements decrease and their surface-to-volume ratio increases, detailed knowledge of the electronic properties of the relevant interfaces becomes increasingly necessary. To analyze our data, we often develop new theoretical models and use them in conjunction with experiments to understand how electrons behave at interfaces.
Angle- and time-resolved two-photon photo-emission has emerged as a technique that directly probes the two-dimensional band structure at an interface and allows the dynamics associated with these states to be studied on a timescale as short as 30 femtoseconds. Two-photon photo-emission is a pump-probe technique which can simultaneously probe initially occupied states and unoccupied states. With a pump laser pulse, an electron is injected from the metal into a thin film layer and a second probe pulse photo-emits the electron. Probing an initially occupied state is possible through a virtual intermediate state. The binding energy is obtained by measuring the kinetic energy of the photo-electron and subtracting the energy of the probe pulse. The pump and probe pulse can be delayed with respect to each other to study the electron dynamics on a femtosecond time scale.
The intermediate states of interest are quite general and include molecular orbital based excited states such as HOMOs, LUMOs, in addition to image potential states (IPS); 2PPE can also probe related interfacial band structure. Image potential states are a series of relatively weakly bound states that exist for all metallic substrates. An image potential state is created by exciting an electron with sufficient energy to escape the metal but insufficient energy for photo-emission. The electron outside the metal surface will induce a polarization in the metal which can be modeled by a positive image charge. The interaction between electron and its image can be solved quantum mechanically, which results in a series of hydrogenic energy levels which converge to the vacuum energy. Moreover, since the expectation value of the electron wavefunction, i.e., the distance away from the metal surface, is only a few Angstroms, an image potential state is a sensitive probe of a layer deposited on the metal surface. Changes in the energy of an image potential state indicate changes in the layer surrounding it.
Interfacial band dispersions (energy as a function of momentum) can be determined by measuring the angle-dependence of the photoelectron kinetic energy. Photo-emission at well-ordered interfaces preserves the electron momentum parallel to the surface (p|| = hk||), given by:
with me being the free-electron mass, Ekin the kinetic energy of the photo-electron, and theta the emission angle with respect to the surface normal. For delocalized electrons behaving like free particles parallel to the interface, the angle-resolved 2PPE data will exhibit a parabolically dispersive band characterized by an effective mass (m*) close to the free-electron value (me):
where E0 is the interfacial band energy at normal emission where k|| = 0.
Alternatively, a spatially localized electron gives rise to a non-dispersive band. In a non-dispersive band, k|| is no longer a good quantum number, and a localized electron can be represented as a superposition of many delocalized parallel momentum state wavepackets. As a result, the probability of photo-emission for a localized state versus angle—and therefore k||—is proportional to the population of the k|| distribution which comprises the localized state. The photo-emission intensity of the localized state versus angle can be used to map out the k|| distribution, and the Fourier transform of the intensity is proportional to the real space distribution of the electron wavefunction, |Y(k||)|2. Within the approximation that the photo-emission final states are plane waves, a direct experimental determination of the spatial extent of electron localization in two dimensions is thus possible.